Skip to main content

Full text of "Theory of interpretation applied to artistic musical performance"

See other formats







































MUSICAL DEVICES AND DETAILS. Prelude, Introduction, Antiphonal Phrases, 
Sequence, Echo, Anticipation, Canonic Imitations, Rhythmic Imitation, 
Parenthesis, Counter-subject, 79 


MUSICAL DEVICES AND DETAILS (Continued). Eingang, Intermezzo, Cadenza, 

Passage, Appendix (Codetta), Refrain, Episode, Thesis, 94 


MUSICAL DEVICES AND DETAILS (Concluded^. Carillon, Ground-base, Drone- 
base, Pedal-note, Recitativo, Coda, Termination, Recollection, Stretto, .... no 


THE DANCE FORM. OLD STYLES. Zarabanda, Corrente, Gavotte, Musette, 

Bourree, Rigaudon, Allemande, 126 





Tarantella, Czardas, Habanera, 132 






MOVEMENT. Prescribed, Variable, Connected ; Movements Influenced by Rhythm, 163 











STYLE AND EXPRESSION. Indications of ad libitum and a tempo, Altering of 
Note-values (Tempo Rubato), Fictitious Note- values, Notation Signs, Addi- 
tion of the Poetic Text, Emotional Expression, 227 


INTERPRETATION IN GENERAL. Style Influenced by Technic ; Melodic, Rhythmic, 

and Harmonic Designs ; Hidden Melodies ; Dynamics Modified ; Dual Themes, 238 


INTERPRETATION IN GENERAL (Concluded). Single and Double Slurs, Delayed At- 
tack, Ritenuto in Forte Passages, Disconnected Motives, Form, ad libitum 
Repetitions, , . 249 






EPOCHS IN Music. The Sixteenth Century, Alte Kammermusik, Altes Clavier- 

musik, Mozart Epoch, Beethoven, Schubert, Nineteenth Century, 272 

INDEX, , 2 g 5 



In writing the present work the author has endeavored to provide 
music students with a tangible and practical manual on Interpre- 
tation. But the limitations of his system must be clearly under- 
stood : Nothing in the nature of positive or inviolable direction is 
intended. And notwithstanding the numerous corroborative illus- 
trations quoted from the highest authorities, the author does not 
claim for any of his deductions the force of a scientific rule. Art 
is not subject to abstract formula. Even the most pointed directions 
are, therefore, merely intended as suggestions to the inexperienced. 

Supplementarily to the various subjects herein treated the student 
should endeavor to acquire individual ideas of style and interpre- 
tation by constant study and the observing of nuance signs in 
standard works. If a section is marked crescendo, endeavor to 
ascertain why the composer gave this direction. Also, why the 
accent ? Why the diminuendo, the slur, or the pause ? Such 
questions are not only pertinent, but important. 

The various shades of meaning which are conveyed by synony- 
mous terms should be carefully discriminated. For instance, rail., 
rit., ritenuto, perdendosi, morendo, allargando, smorz., etc. Also, 
con anima, con moto, con brio, con fuoco, strepitoso, stringendo, piu 
vivo, piu mosso, stretta. 

Marked accentuation usually applies to marches, to most of the 
dances, and to music of a grand, majestic, or heroic character. 
The legato style is inclined to seriousness ; the staccato to lighter 

In certain works punctuation is essential (Serenade in D-flat, 
Borodin); in others no points are to be observed, excepting at the 
close of periods, Clara Schumann, op. 21, II. 

The recent treatises upon piano pedals by Venino, Schmitt, and 



Kunkel have saved the author from extended comment concerning 
these important mechanical attachments. Students are therefore 
referred to the works mentioned, and to such special editions as those 
by Sherwood, Foote, Bohlmann, and W. G. Smith, in which the most 
approved method for using the damper pedal is plainly indicated. 

Particular attention is directed to von Biilow's analytical editions 
of works by Haendel, Beethoven, and other masters. These 
editions comprise one of the most important contributions to the 
small store of reliable musical didactics. Von Biilow's annotations 
of the Beethoven piano sonatas extend from op. 53 to in. 

To assist the student in making theoretical analyses of music 
selected for study the author has devised a series of abbreviations. 
These, with the corresponding key, may be found immediately after 
the Index. 




Let it be supposed that a given composition contains no perpen- 
dicular bars, and that no mensural signature appears. If the 
music be of a preludatory character, or consists of fragmentary 
groups, the absence of equal time-divisions, called measures, might 
be proper. Indeed there are passages whose interpretation would 
be facilitated by such omission. For example, the final cadenza in 
Godard's " Guirandes." 

But as a general rule the absence of mensural divisions would 
result in befogging the outline and confusing the delineation. Such 
an instance is here presented : 

Ex. 1. 

J. S. Bach. 


The effect is analogous to that produced by the reading of a com- 
pound sentence without accent or pause. 

The necessity for mensural divisions may, therefore, be consid- 
ered as a fundamental law in nearly all kinds of music. The few 
exceptions apply only to peculiar and isolated instances, such as 
the last of the adagio in P. E. Bach's F-minor Sonata. 



The rhythmic peculiarities of various kinds of measure claim 
first attention. These are indicated in a general way by the men- 

jQ_ ^. ^4. _*!. -t* O 

sural signature, such as ^g , *, ^T, *, If, 4t, and so on. 

"4r ^E ~%E o "Cf o 

Since each measure represents a regular time-division of music 
there must be some means of disclosing and marking these divi- 
sions. A regularly recurring accent (as the books tell us) serves 
this purpose, a proposition that needs no demonstration. 

Objection has been raised against the modern application of 
"time signatures" because they do not sufficiently indicate, or 
correspond to, the motives and phrases of certain compositions. 
It would, however, require an immense variety of combinational 
figures to accomplish this result, and the composer would be 
obliged to continually alter the time signature. After all the 
bother of this complicated system it is doubtful if our compre- 
hension of rhythmic designs would be greatly aided. 

The distinctive features of various kinds of measure are impor- 
tant considerations, though they have generally been overlooked. 

A brief statement of the author's theory concerning these points 
is contained in the following paragraphs. 

What principally concerns us here is, not the number, but the 

kind of notes indicated by the mensural signature. If this be ^, 


the two quarter-notes fall upon regular beats, the first being more 
prominent. Therefore, the following order will prevail : 


This would plainly indicate two regular beats 
mp p mp p 

in each measure. Now, suppose the rhythmic arrangement were 
changed to four eighths. The beats remain the same, and conse- 
quently the second and fourth eighth-notes would be sounded more 

softly than the first and third eighth-notes : 



With notes of lesser value the same principle applies, thus : * 

* A dynamic consideration naturally enters here, but this must be explained later. 


I I 

The signature of common measure indicates four beats, without 
regard to the rhythmic contents. The composer may use a whole 
note as a measure unit ; two half-notes ; four quarters, or any 

other equivalent of -: the four quarter-beats still remain. The 
first I is most prominent because it marks the beginning of a 
measure. The third ! is relatively next in importance because it 
marks an equal division of the measure. The second j comes 

next, and the fourth I is the least prominent. These (2 and 4) 

must not be ignored, as they fall upon the regular beats. If the 
composer desires every other tone to be unaccented he employs 

some such arrangement as this : ******** 

Q K 

We are thus led to a comparison between ^ and ^F measure. 

A distinction is to be observed between 

n n 



, supposing that both are executed in the same 

space of time. The second and fourth eighth-notes of the first 
illustration receive no mark of emphasis ; whereas the second and 
fourth quarter-notes fall upon regular beats and must therefore be 
duly marked. This theory applies directly to such instances as 
the Turkish march from Beethoven's " Ruins of Athens " : 

PP > 

The only distinct accent here falls upon the beginning of each 
group, excepting in the fourth measure,, where the sustained tones 


(d and f) are slightly marked. Surely it cannot in truth be said 
that the effect would be identical in common measure, thus : 

If correctly performed this results in a perversion of the composer's 
design, and degrades it to the level of a common parade march ; 
whereas Beethoven's conception is purely ideal. That famous Alle- 
gretto in the master's Eighth Symphony affords a still stronger 
argument in favor of these distinctions between greater and lesser 
note-values. A fragment is quoted : 

Ex. 4. 


By means 01 the slight accentuation a very dainty, fairy-like 
effect is produced. But if we lengthen the value of the notes, and 
increase the quarter-note movement accordingly, the characteristic 
effect which Beethoven intended would almost entirely disappear. 

The same distinctions are to be observed between gj and *f 

measure, the former being lighter and more graceful in character. 
The following instances are cited : Weber, Romance in F t (E) ; 
Mendelssohn, op. 72, VI, (E) ; Clara Schumann, op. 21, II, (E) ; 
Jos. Low, op. 485,11 and VI, duets, (E) ; R. Schumann, "Am 
Springbrunnen," from op. 85, (M).* 

In the little hide-and-seek piece from the same opus Schumann 

employs V^r measure in order to still farther reduce the amount 

1 U 
of accent. See " Versteckens." 

*The gradings (E, easy ; M,. medium ; D, difficult) are merely approximate. 



Also there is a difference between six eighths in measure 

and two triplets in 

measure : 


Only the first tone of each triplet is marked ; 


whereas every eighth in J [ measure receives at least a slight 

accent. In slow movements these distinctions are particularly de- 
sirable ; in fast movements it is not always possible nor essential 
to observe in the performance such minute details of measure and 
rhythm, excepting in such light and delicate movements as Schu- 
mann's " Versteckens," previously mentioned. 

In applying these precepts to future illustrations it will be neces- 
sary to employ the symbols which represent three species of accent. 
These are : (i) A, (2) >, (3) . The more forcible sf (or/s) is an 
unusual accent, and will not be required in present examples. The 
arbitrary application of these mensural accents primary, second- 

ary, and tertiary is as follows : 





until they can be performed accurately. 

These should be practiced 

The second beat in ^T measure might, under certain circum- 

stances, require a secondary accent. Where the contents of the 
measure naturally fall into two parts this plan would be preferable. 
For example, in the following : 

Ex. 5. (a). 



>g IT: aim -Tra=r 

* -h* A T- A A > 

-J ^--F- -1-aH-H* I-^H 

Jf A_JJ ,.M I l_jg_i_^^ 1^^^ . I 


"Oriental Pictures," IV. Schumann. 



A > 

Similar examples may be treated in this manner ; but in the 
majority of instances the tertiary accent upon the second beat would 
be preferable. See Ex. 2. 

One more consideration is involved in the treatment of this sub- 
ject : So far as mensural accents may be considered they naturally 
diminish from the initial impact of each succeeding measure, as 


shown by the dynamic symbol : 

This is not to be interpreted literally. It is in- 
tended to illustrate merely this : that the force of accent recedes 
from the initial point, and that the minimum degree of stress 
occurs where it is farthest removed from the beginning of a meas- 
ure. Therefore, primary, secondary, and tertiary accents become 
necessary in expressing the fundamental features of rhythmic 

Additional selections for mensural accent : Berceuse, L. Schytte, 
op. 23, VII, (M) ; " Bagpipes," H. Ryder, (E) ; Waltz, op. 38, 
VII, Grieg, (M). Light mensural accents may be applied to the 
left-hand part throughout. 




Mensural Accent. A wider application of accent is now to be 
considered. The student must understand at the outset that the 
formula of mensural accentuation as described in Chapter I has 
but one fundamental object, and that is to indicate the time -divisions 
called measures. This should be constantly borne in mind. If the 
melody is of such character as clearly to define the measure and 
rhythm and movement, then there will be no occasion for mensural 
accents in the accompaniment. But when the melody does not in- 
dicate the regular measure, the accompaniment must supply this 
deficiency by means of periodic emphasis. 

The conditions under which the formula of mensural accentu- 
ation may properly be enforced lie, therefore, at the foundation of 
our endeavors at musical interpretation. 

One of these conditions appears in the preliminary matter which 
frequently precedes a principal theme, and by means of which the 
composer intended to indicate the key, measure, and elemental 
rhythm of the accompaniment. Thus, from R. Volkmann's 
" Picture Book," op. i.i : 

41 In the Mill." 


r Q , * 

Ex. 6. 

i i i i ' i i- 1 I i i i i i i 

*--*--*-H **-*- -*-* +--*+-** 

These two measures of prelude should establish certain points : 


(i) That the measure is ^g- ; (2) that the movement is moderate; 


(3) that the accompaniment has a uniform rhythm of sixteenth- 
notes. This is the groundwork. 




The primary and tertiary accent marks here apply very directly 
and properly. 

When we come to the theme it will be necessary to modify this 
arbitrary mensural emphasis. 

Another preludatory example is given : 

Bolero. Moszkowski. 

Kx. 7. 


Key, measure, and movement are here indicated ; also the char- 
acteristic castanet rhythm of the bolero. The second measure 
is to receive the same accentuation as the first, in order clearly to 
define the triple-beat measure. 

After these preliminary objects have been accomplished the 
regular emphasis marks may be omitted especially when their 
continuance would prove either monotonous or obtrusive. 

Intermezzo, passage, and termination present additional instances 
in which the regular beats require a distinguishing mark. This is 
more especially true when the notes are rapid, or where no distinct 
melodic outline appears to claim attention. A simple intermezzo is 
quoted from F. Kuhlau : 

Ex. 8. 



It will be observed that only two primary accents are here in- 
cluded ; but the measure had previously been determined, and 
therefore it is not necessary to mark the beginning of each measure 
in the same manner. The secondary accent should be sufficiently 
pronounced to prevent any doubt as to the rhythmic outline. A 
slight accent is placed over the initial note of each group to indi- 
cate its beginning, though according to our formula this part of the 
measure is unaccented. (Such instances will be illustrated more 
fully in Chapter IV.) Ex. 8, from the rondo in op. 20, I, should 
be thoroughly tested by omitting the accents, and then performing 
it exactly as written. 

The next quotation is a better illustration of mensural accent- 
uation : 

Ex. 9. C Major Sonata, XV. Mozart. 




b l- 

This is in form of a passage, the design of the first measure 
being continued in sequence. All such instances demand metrical 

These simple extracts illustrate an important fundamental principle 



in artistic interpretation, and this principle should be practically 
applied before proceeding farther. 

Rhythmic Accent, Motive, Phrase. Motive has become a 
very flexible term in musical literature. In Wagner's music-dramas 
the typical motives (material, character, sentiment, and phenomena) 
vary in length from one to seven measures, though generally 
they correspond to the phrase of two measures. The two first 
tones of the Sword figure may be considered a motive, since these 

Ex. 10. f 

would appear prominently in a development of the phrase. But 
the complete Sword motive is this : 

Ex. 11. 

This phrase is complete in itself; it is all that Wagner chose to 
write for this material suggestion. But it is frequently advisable 
to adopt a more technical view by considering as a motive any 
group or figure which is sufficiently suggestive to admit of melodic, 
harmonic, or rhythmic elaboration. According to this view, the 
first phrase of Beethoven's op. 2, I, contains two motives (or semi- 
phrases) ; thus, (a) and (b) : 

Ex. 12. 

In the first and second subjects, as well as in the development, 
the composer treated these as separate motives. They constitute 
the two phases of the subject the contrasting features. One is a 
motive quite as much as is the other, since they are undoubtedly 
different in aspect and in effect. 

The simplest examples of phrase contain two measures, as in 
Examples n and 12. In lyric music the phrase forms a natural 


subdivision of the period. If the phrase does not form a sub- 
division, it is not to be treated as such. 

There are three constituent elements to be considered in analyz- 
ing the phrase : melody, rhythm, proportion. A phrase from 
Haydn is selected for analysis : 

Ex. 13. 

(i) Melodically this consists of a simple tonic chord figure ascend- 
ing and descending. This may, like most phrases, be divided into 
two motives, the ascending and the descending. (2) The rhythmic 

,773 I jj j (3)Themensu- 

arrangement is primitive 

ral proportion is quite equal two measures. This is the usual 
length of phrases. 

The student must be sufficiently familiar with the melodic and 
rhythmic features of a given phrase to be able to trace it through 
the various processes of development and metamorphosis to which 
it is susceptible. For this purpose it would be well to examine 
closely this andante movement from the " Surprise " symphony. 
The treatment is both simple and ingenious. 

As the measure is indicated by means of a principal accent at its 
commencement, so must the advent of the phrase be indicated by a 
rhythmic accent. And since the measure is only part of the phrase, it 
follows that the former must be subject to the latter. If the mensural 
accents are successively maintained the effect will be to divide the 

phrase into two semi-phrases, thus : 


* J ' 

the music does not justify this subdivision. The phrase should be 
recognized as a unit (even when it contains two contrasting features), 
and therefore the primary accent is to be reserved for the commence- 
ment of the phrase : 

1 \ rlg^ 

Ex. 14. 



This is rhythmic accentuation. The second phrase is to be 
marked in the same manner : 

Whether the style be fast or slow, loud or soft, legato or staccato, 
the principles of rhythmic accentuation remain the same. In the 
present instance only a slight emphasis is required to outline the 
phrases. As a preliminary study it would be well to perform one 
or two periods in the manner indicated by Example 14. But in 
strict designation these simple, natural melodies do not require a 
rigid application of theoretical formulas. 

Another two-measure phrase is selected : 


Lar ghetto. 

Ex. 16. 

This is more smooth and flowing, and would require special 
treatment. But as far as mensural and rhythmic accents are con- 
sidered, the application is very simple. 

The next quotation illustrates a different phase of this subject. 
On account of the long-sustained tones in the treble it becomes 
necessary to apply mensural accentuation in the accompaniment 
below : 


Ex. 17. 


|= EgfE 



The dotted half-notes in Example 17 could not, at the beginning of 
a movement, indicate the kind of measure employed. Mensural 
accents are therefore essential. Also, when the theme is figurated 
or contains arabesk work, it becomes necessary to plainly outline 
the measure and the rhythm, thus : 

Ex. 18. 

E. Nevin. x Op. 13, V. 

The seeming peculiarity in rhythmic arrangement was intended 
to indicate the notes for each hand ; in effect the order is perfectly 
regular : 

Example 18 is the initial phrase of this barcarolle, and the 
listener's impression as to measure, rhythm, and movement remains 
indeterminate until these features shall have been fairly revealed by 
the performer. Hence the regular accents are more essential here 
than they are in the repetition of the phrase which immediately 

A number of additional illustrations are mentioned in connec- 
tion with this chapter. 

Moszkowski : "Spanish Dances," op. 12, prelude to No. I. 
Mensural accents. Preludes to III and IV, the same. Following 
the preludes, rhythmic accents take precedence. These dances are 
of medium difficulty. (Two or four hands.) 



Mendelssohn's " Songs without Words " : Prelude to I ; men- 
sural accents. Also the following : IV, introduction ; VI, prelude ; 
VII, IX, X, XII, XIII, XV (six measures), XVI, XVIII (very 
little accent in the prelude, and still less in the accompaniment after 
the rhythmic accent begins), XX, XXI, XXIII, XXIV, XXIX, 

Nearly all these instances are preludatory, and require mensural 
accentuation until the theme begins. The emphasis, however, must 
be light and not attract too much attention. 

With exception of the preludes and intermezzi, rhythmic accent 
is to be applied. 


The common definition of Slur is, a curved line drawn over two 
or three notes to indicate that they should be played smoothly. Doctors 
H. Riemann and Carl Fuchs, in their Guide to the Art of Phrasing, 
give a different view : " The employment of the slur as a sign for 
legato is quite abandoned (sic), inasmuch, at least, as its presence 
neither necessitates legato nor excludes staccato. However, it is 
assumed that notes included under a slur are to be played legato, 
whenever staccato marks do not call for the contrary. . . . 
Now, the new office of the slur is to indicate the 
articulation of the musical thought into its natural divisions 

The contradictions will be noticed, and indeed it is a contradictory 
subject. But the music itself must speak to us and aid us in the 
solution of all these problems. 

Here, as in other instances, it seems desirable to present the two 
phases of the subject separately : a practical, elementary illus- 
tration of the slur here, and a broader application in the latter 
part of the book. This will obviate the necessity for even seeming 
contradiction, which is always puzzling, if not detrimental, to the 

The slur is a symbol of unity. It is not restricted as to length, 
but may embrace any number of notes. This fact tends to prove 
that the slur is intended, primarily, to signify that the notes encom- 
passed by it form a division or subdivision of the music. The 
short slurs in Moszkowski's " Air de Ballet " (second period) in- 
dicate motive figures ; the long slur in Saint-Saens, op. 24, II, 
represents ascending and descending groups, which the composer 
wishes joined into a section. This signification the slur always has. 
Whether the notes beneath it are to be played legato or non-legato, 



depends upon the character of the music ; though usually the slur 
is a sign of connection as well as unity. Where the only intention 
is to indicate a phrase or a group, the bracket should be used, thus 
confining the slur to its original import of showing the number of 
notes to be connected, and where the connection ceases. Such 
will be the application in the major portion of this system. 

An excellent illustration, admitting no doubt as to the meaning, 
is here quoted from A. Loeschhorn, op. 101, I : 


This slur embraces the first phrase, which is to be played legato. 
C, at the end of the slur, is to be slightly disconnected. After 
repeating this phrase the next is divided into two semi-phrases. 
Hence the two separate slurs, as ptr Ex. b : 


The notes within the first slur form an independent group ; those 
within the second slur are a repetition in sequence of the first. 
The two groups are therefore separated, as indicated by staccato 
marks. The latter merely tend to relieve any doubt which might 
exist as to the last note of these slurs being disconnected. 

The last phrase is subdivided into demi -semi-phrases, and these 
are slurred separately : 


The melodic construction as well as the harmonic basis of these 
short motive groups demand a separating of one from the other in 
order to impress their individuality upon the listener. It should 
be observed that the last note of these slurred groups is in every 
instance a note of small value, naturally short. Therefore, in order 
to separate the groups the performer must employ a staccato touch 
at the end of each group where the slurs end. 

Assuming that the slur indicates legato, it is argued (and has 
been published as a rule) that the end of a slur is a sign of staccato. 
This merely corroborates the old adage that " a little learning is 
dangerous." We may say that the termination of the slur shows 
the limit of connection, for so it does. But if the last note is long, 
we are not privileged to make it short. (This will be duly illus- 

And there is another " rule of phrasing" which needs consider- 
able qualifying. It is this : That " the end of a slurred group or 
passage should be light.". But if the terminating note falls upon 
the accented part of a measure, then the rule is invalidated and 
worthless. These points, as well as that other one in regard to 
the accented beginning, will be here illustrated. 

A brief example is quoted from the finale to Mendelssohn's 
" Italian Symphony " : 

Ex. 21. 



l i 



T- t t 


*\ s s 



v n 



t K 


1 ^ 





-* H 

ft j r 



. & t~ 





* 1 



p P 




^ 1 



i r 


1 r 





~ .. 


The rhythmic groups are perfectly symmetrical, but they begin 
upon 2 and end upon i. The slurs call attention to this fact. If 
the rhythmic arrangement can be impressed upon the listener by 



means of ordinary accentuation, then there will be no occasion for 
syncopated accents. The first group is isolated from what pre- 
cedes, and since the passage is forte the beginning of the rhythmic 
group will naturally be marked. The last note at the end of each slur 
has so little time -value that its proper performance is equivalent to 
staccato. But this does not interfere with the mensural accent on the 
first note of the measure. In such instances the following interpre- 

tation may be adopted : (b). hjjf"" j^*^ 1 

The last triplet, according to the dynamic symbol, is to be played 

thus : (c). - 

When performed by an orchestra, the 

ff f 

second half of each measure is marked by the trumpets, and thus 
the mensural equilibrium is maintained even if the second beat 
should be unduly emphasized. See Ex. 21 (a). 
Following this passage two larger groups appear : 


These rhythmic groups (a) and (b) correspond, and a slight 
separating of the tones serves to distinguish one from the other. 
Here, as before, the last note of group (a) is unaccented, not 
because it terminates the slur, but because the last note of a 
triplet receives no emphasis. At the end of the second slur the 
composer marked the tutti chord ff. 

Primarily, the slur itself has nothing to do with accent. Here, 
for instance, the slur means legato : 

J. Raff, 
(a). > (b). 

Ex. 23. 

s^ii 1 ? 


The notes above the slur at (a) are to be connected, and at (b) they 
are to be disconnected. But the curved line has no dynamic 
significance whatever. Observe, also, that the notes at (a) are 
unaccented, though the slur begins here. The accent following 
(indicated by the composer) is quite natural, since it occurs upon 
one of the regular beats. 

The beginning of a motive or group may, in certain situations, 
require an accent ; but it will not be so on account of the slur. 
The commencement of the slur merely shows the commencement 
of the group ; if an accent is contemplated at that point, we must 
ascertain the cause from some other source, surely not from the 
curved line above or below the notes. 

In the next example the slurs direct attention to the symmetry 
of the melodic groups : 

C. Reinecke. Op. 129, II. 

Ex. 24. 

In performing this the pianist must determine whether the groups 
are to be indicated to the listener by means of disconnection or 
accent. The slurs will not afford the least assistance in arriving at 
a proper conclusion. The whole phrase might be slurred or the 



curved line omitted altogether ; the actual groups would still 
remain to be considered. 

The careless manner in which slurs are sometimes placed fre- 
quently destroys the unity of a melodic figure by dividing it into 
two or more parts. This is the most serious objection to certain 
"methods of phrasing" : They sacrifice the greater to the lesser 
effect. For instance, here : 

Mendelssohn. Op. 19, II. 

N * *FH- 

Ex. 25. p-XU-^P- ~p^ZI^ZI^j ~| j**4 0^^~l[ (Riemann's phrasing.) 

There is no good reason either in technics or esthetics for such 
hysterical interpretation. 

Errors like the following are of common occurrence, though 
some of these may have been perpetrated by the engraver : 

^^ Merkel. Op. 126, I. 

Ex. 26. 

Evidently these slurs should have extended a little farther, since 
the scale passage ends on C, not on D. From the same opus a 
similar instance is taken : 

Ex. 27. (a). 


BT^EIE^^Fi^j^E^F^ 11 : FraE3*= 


"^^ * 


2 9 

The first scale passage leads naturally to F, and there should be 
no disconnection before this end is reached. So with the next 
scale, which leads to C. In all such instances the following 
reading is recommended : 


i i 

i i 

There are many examples (otherwise correct) which have a 
tendency to mislead the inexperienced in their attempts to apply 
the slur. For instance, this extract from a tarantella : 

F. Thome. Op. 43. 

The composer did not intend to indicate a separating of each 
group ; yet certain performers would so construe the short slurs. 
These merely show that each hand executes a triplet group. But 
the upturned and downturned stems are sufficient indications of the 
fact that both hands are employed alternately, and the design 
would be more clearly represented in this manner : 


Each measure here represents a figure which is continued in 
sequence. The bracket (or a long slur above) continues until the 
figure changes, and represents about one-half of the entire cadenza. 


An instance somewhat similar may be found at the close of Mr. 
Emil Liebling's Canzonetta, op. 26 ; but there the entire sixteen 
gruppetti are very properly slurred together as a series of short 
figures united. 

The principal difficulty in applying the slur will consist in deter- 
mining whether it was intended to indicate the connection and 
disconnection of tones, or merely to call attention to melodic or 
rhythmic divisions and subdivisions. 

In connection with this lesson the following may be consulted ; 

Sonatina in D, Isidor Seiss, op. 8, I ; Sonatinas, G. Merkel, op. 
126, I and II ; Sonatina in C, A. Loeschhorn, op. 101, I ; Sonatina 
in D, J. Handrock (Th. Bohlmann's edition) ; Idylle, Wm. H. 
Sherwood, op. 5, II. 




In this system a very broad distinction is made between Accent 
and Punctuation. Under different conditions both are employed 
to the same end i. e. y the exposition of musical periods and their 

That which pertains most directly to the organic construction of 
music, whether it be formal or fantastic, is indicated or expressed 
by rhythmic accent or by punctuation. This objective view must 
constantly be borne in mind, since these primary elements of 
expression are frequently forced beyond their legitimate sphere of 
action into the intangible realms of emotion and fancy. 

Punctuation is here to be understood in its literal sense. In the 
majority of instances it will be employed to point the melodic or 
rhythmic divisions. Also, it will serve to modify the rhythmic 
accents ; and frequently the punctuating mark may entirely obviate 
the necessity for emphasis when the latter quality is not desirable. 

But the author would not like to affirm the statement, so often 
made during recent years, that the majority of phrases are to be 
separated from each other. We must, of course, determine upon 
the length of the phrases whenever the music contains these sub- 
divisions ; but we must also consider whether short punctuations 
are desirable or undesirable in certain instances. This will be 
illustrated in what follows. 

As a preliminary demonstration let it be supposed that a com- 
plete period is selected for performance, and that this period con- 
tains four equal subdivisions, commonly called phrases. If each 
phrase is really a subdivision, if it corresponds to the metrical line 
of a poetic stanza, then a punctuation like the comma is to be 
expressed at these points, as in the following : 




" In dreams I walk in pleasant ways, 
By limpid streams in sunny dells, 
Where peace abides and' beauty dwells, 
And splendors glow through happy days." 

A corresponding musical period is now presented : 

Ex. 29. Cabaletta. Th. Lack. Op. 83. 

-. _r ----, 

1 \*-* H-^-*-^-t I H 

This, like the poetic stanza, falls naturally into equal subdivisions ; 
so much so that very little care need be bestowed upon the phrase 
points. The commas are included merely to show the structural 
features of this period. With exception of the fermata the punc- 
tuations are not absolutely essential, because the style is mostly 
demi-staccato. Rhythmic accent is therefore much more important 
here, particularly at the beginning of the third phrase, after the 
staccato scale figure. 

When the period has come to a satisfying close by means of a 
complete or perfect cadence, a brief pause is usually to be made 
upon the final tonic to represent a sense of completeness ,or repose. 

Another simple period, somewhat different in style, is quoted : 

Ex. 30. 


Allegretto, Seventh Symphony. Beethoven. 

1 1 





These short phrases are marked by the characteristic rhythm, 
which, by its uniformity, tends to reveal the 


outlines without the aid of further punctuation. 

Phrases Beginning After the First Beat. It may be stated 
as a general rule that phrases which begin upon 2 will end upon I ; 
those which begin upon 3 will end upon 2, and so on. This will 
apply to triple or quadruple measure. Such instances more fre- 
quently require some kind of punctuation than do those which 
begin upon the first of a measure. The latter are sufficiently indi- 
cated by means of the regular accentuation, as exemplified in 
Chapter II. Examples . 20 and 22 afford further proof of this 

The rule that " beginnings of phrases should be accented " must 
be applied with cautious discrimination when these beginnings fall 
upon an unaccented part of a measure. We might, with as much 
reason, say to a reader, " the commencement of every sentence or 
clause must be accented." Whenever it seems desirable to apply 
the formula to music, the performer must have a care that the 
mensural equilibrium is not too greatly disturbed thereby. This 
will be illustrated in what follows. 

Two four-measure phrases, beginning upon the last beat, are 
here quoted : 

Ex. 31. (a). 

Moszkowski. Op. 12, III. 



|_ m 


p~z jp 

m * m * 

-M* 11 ; 

- -i 

By raising the finger gently from the C ft key on the second beat, 
immediately before b, the punctuation will be sufficiently expressed. 



The effect will be like this : 


second phrase, ending on B, is to be treated in the same manner. 
The beginnings of these phrases require no accent to mark their 
advent. These two phrases are isolated and there can be no 
reasonable doubt as to their beginnings ; besides, they are marked 

Farther on, where the phrase begins on the second eighth note, 
the composer distinctly reserves the accent for the first of the fol- 
lowing measure : 

Ex. 32. nft^^A H*. * 

The mensural accents here are retained for all the semi-phrases, 
which begin staccato. Before taking leave of Ex. 3 1 it should be 
stated that sufficient accent is supplied by the accompaniment, 
which clearly defines the measure and movement. 

The next eight measures are joined together by the composer 
into a section, and this requires special treatment : 

1 . ^~~l 


No punctuation is advisable before the close of the section ; in 
truth, it is distinctly forbidden by the composer, who is supreme 
judge in all such matters. But for the sake of symmetry we 
emphasize the first tone of the last two groups, A. While it is true 


that the separate phrases begin upon the last of a measure, it is 
also true that this feature disappears during the united section, and 
the following groups begin with the first of each measure. There- 
fore the normal accentuation is not disturbed, though this must be 
very light. The symmetrical construction of this charmingly 
characteristic dance, and the fact that the first sixteen measures are 
pianissimo, forbid the use of marked accentuation. In the repeti- 
tion of this period the divisions and subdivisions are identical : two 
phrases of four measures each, and one united section of eight 
measures. Therefore the accents and punctuations occur in the 
same (corresponding) places. 

The appearance of semi-phrases usually calls for some form of 
punctuation, because these small subdivisions are proof of the fact 
that the phrases contain two motives and are therefore of a dual 
character, thus : 

Andante. * ' Narcissus. " E. Nevin. 

Ex. 34. 

t^p-^^t Fp t^L i-bH = H-i-4&g- - 

Phrase. ^ 

This phrase contains two semi-phrases. The separate slurs are 
to be considered as negative rather than as positive signs of dis- 
connection. The melodic flow of sound must not be interrupted, 
and yet the two phases of the subject should be revealed. A non- 
legato style at | would be expressive and proper. The end of the 
phrase may be punctuated as with a comma, which is to be under- 
stood as more pointed than the non-legato sign, | . 

A peculiarity to be particularly noted occurs in the third phrase, 
thus : 

_L0 -_* -- ^-T 

The principal aim should be to join the little rhythmic groups 
together, a la portamento, rather than to separate one from another. 
Perhaps the following indications would best represent the mood : 


> > 
This tends to reveal more plainly the melodic outline : 

In every instance the first of the slurred couplets, F, is to be briefly 
sustained after the following key has been pressed down. This 
contributes, noticeably, to that peculiar quasi portamento effect 
which a singer or violinist would impart to this graceful and orig- 
inal theme. A gentle hand pressure from the wrist is employed in 
all such instances. No perceptible disconnection of the tones 
occurs until the end of the phrase (,) has been reached. 

Much more pointed are the following demi-semi-phrases from 
M. Moszkowski's op. 12, I : 

Ex. 36. 

Staccato marks are included in addition to the short si urs ; there- 
fore the rhythmic groups are considerably isolated. The character 
of the music, however, is quite different from that of Ex. 34. 

Observe that the triple-beat measure temporarily disappears, but 
is restored at the end of the phrase. The actual effect is like this : 



These metro-rhythmic contrasts tend to relieve the monotony of 
a uniform succession of beats, especially when the movement is not 
interfered with. 

There is another rule (supplemental to the one regarding initial 
accent) that the end of a phrase or slur must be light. But the 
direction regarding the phrase applies to isolated instances only. 
There is but one example (second phrase of 29) in this chapter to 
which the rule would properly apply. Where the last note of a 
phrase or group is of brief duration and occurs on the weak part of 
a measure, then the last note would be light. Hence the first four 
slurred groups in Ex. 36 (b) come within the jurisdiction of the rule. 
But the end of the phrase falls upon the first beat and is, therefore, 

Furthermore, if this last note were immediately followed by an- 
other phrase, and if the two phrases required punctuating, then the 
last note under the slur would be played staccato. This is illus- 
trated here : 

Ex. 37. (a). 


In order to separate these phrases one from another without 
retarding the movement, it is necessary to employ a light staccato 
touch at the points indicated by commas, thus : 


And still more so here: ( c ). II 

All these terminations fall upon the second beat, which receives 
but a slight emphasis even according to strict mensural accent- 
uation. The end of the period is expressed by means of a brief 
pause while the tones diminish. Other instances, where the phrase 
terminations should be light and short, will be specified as they 
occur. But it is not advisable to deduce from these simple illus- 
trations any guiding rules which would prove arbitrary and con- 
flicting unless they were applied merely to parallel instances. We 
must, therefore, examine a great variety of examples in order to 
determine, eventually, how far a systematic mode of procedure may 
be carried. 

Certain phrases are sufficiently punctuated by their manner of 
representation. For instance, here : 

Ex. 38. 

Allegro. Nevin. Op. 13, III. 

The eighth rest at the beginning of the second phrase, b, signi- 
fies silence, and this is a sufficient punctuation. The other phrases 
are similar. 

This principle is illustrated differently in the next quotation : 
Ex. 39. Grieg. 

| , , , 

I 9 I 

* -^V 



Also in the same composer's op. 3, III. In these instances the 
punctuating marks (rests) are supplied by the composer, and the 
pianist need not further concern himself on this point. But, un- 
fortunately, creative artists are not always so particular in noting 
their scores, and hence there are many compositions in which 
necessary punctuating marks are not indicated. These we must 
supply. It also happens that phrase points are frequently pre- 
scribed by well-meaning but misguided annotators, where the 
composer did not intend any such interruption or disconnection. 
This seems to the author a more grievous error than to omit the 
punctuation altogether. 

Attempts have been made to apply a semicolon at the end of 
sections ; but this is unnecessary and unwarranted. Mattheson 
put forth this notion in the year 1737, and as a first attempt at 
practical analysis it was ingenious and plausible. But the actual 
instances to which this rule might apply are so exceptional as to 
render it valueless. 

" The Humoreske," by Grieg, contains a peculiarity to be noted, 
though the period construction is regular : 

Ex. 40. Opus 6, I. 

At the end of the second phrase the last note is tied, and thus 
joined to the following, c. This necessarily excludes a punctuat- 
ing mark and unites the two phrases, as shown by the connecting 
brackets and the tie. The main accent, therefore, falls to the 
accompaniment at c, as the secondary accents did in Ex. 17. 

This manner of integrating two united phrases (partially exem- 
plified in Ex. 33) maybe compared to the longer sentence followed 
by the shorter one in this stanza : 


"This morn there was frost on the meadow, 
The trees are all shivering with fear, 
The grass that was green on the hillside 
Is dying, and with it the year." 

The metrical arrangement is perfectly regular, but by means of 
a slightly extended sentence, from the third into the fourth line, 
the monotonous, rhyming cadence (so prevalent in verse of this 
kind) is avoided without marring either the sense or the euphony 
of the stanza. The conceit applies directly to musical phraseology. 


Rondo, op. 47, I, Reinecke. In the principal theme punctuations 
are required at the end of sections (eight measures), but not for the 
first and third phrases. When the style becomes staccato, accent 
must be substituted for punctuation. Grieg, op. 6, I, rhythmic 
accentuation throughout ; no pause at the end of periods. Op. 6, 
III, to be punctuated regularly. Op. 12, II, four-measure phrases ; 
more accent than punctuation. Op. 12, III, punctuation indicated 
by rests. Op. 12, V, only the sections and periods are to be punc- 
tuated ; accent must do the rest. Brief pauses apply to the periods 
ending in F-sharp minor. Op. 12, VI, the phrases are so well 
defined that no pointed punctuations are required to separate them. 
Non-legato will apply here. Pause before beginning the second 
subject in D-minor. Op. 12, VII, the periodizing is regular. 
These pieces are in the easy and medium grades. 



Following is the author's enumeration of the various methods 
and means of musical exposition as herein applied : i. Mensural 
Accent. 2. Rhythmic Accent. 3. Staccato. 4. Demistaccato. 
5. The Rest. 6. Diminuendo. 7. Rallentando. 8. Ritenuto, or 
Lento. 9. A Tempo. 10. The Fermata. (Tone-quality, and 
parentheses also might be included, but they are treated in a sepa- 
rate manner subsequently.) 

The first five of these points have been sufficiently explained for 
present purposes. The others will appear more plainly as we 

Where the last note at the end of a phrase is of considerable 
duration (as in the G -minor Gavotte by Bach) staccato can not be 
applied to it without sacrificing too much of the note -value. In 
such instances about one-fourth is subtracted from the value of the 
half-note as a means of punctuating the phrase, thus : 

Ex. 41. 

The actual effect is noted in the lower staff*. Since the tones are 
not to be connected beyond that point where the slur terminates, it 
is evident that the last tone must be separated from what follows. 


The interpolated eighth rest serves this purpose.* A staccato 
effect would not be proper in such instances, though it would in 

the following : 

X. Scharwenka. 

(a). "^ " ' 

The end of the slur in these motive figures is plainly intended to 
be staccato, as though written like (b) or (c) : 

These various modes of representation are usually synonymous 
in allegro movements, though considerable experience is required 
in order to properly apply them. 

The next illustration is similar to Ex. 41 : 

Ex.43. "AmMeer." Schubert-Liszfc. 

The first phrase is to be very slightly punctuated at | . As the 
movement is slow, it will be sufficient to take ^ from the value of 
the dotted quarter-note. At the end of the second phrase, b, % 
of the ) 9 may be subtracted by including an eighth rest. These 
punctuations are here illustrated : 

^ ^^ B_ 


* The length or shortness of the rest depends somewhat upon the action of the 



Care must be exercised in such instances not to approximate a 
staccato effect. 

Music of a graceful, tender character requires that the terminat- 
ing tones be quitted rather reluctantly, with but little disconnection 
in the midst of a period. A simple example is quoted from F. 
Kuhlau : 

Ex.45, (a). 


The second phrase, b, is subdivided into short motive groups, each 
of which requires .a brief punctuating mark. At the end of the 
first slur the hand is to be gently raised exactly as the fifth eighth 
is counted. The other punctuations come upon the second and 
fifth counts of the measure. In these instances a small fraction is 
taken from the value of the punctuated notes. But if a rest follows, 
as at c, the note is to be given its full value and the hand is not 
raised until the sixth count. Both points are illustrated here : 

1 2 3 

4 5 

123 456 

123 456 

The mark | shows where the hand is to be raised, but this must 
not be done abruptly, for only a slight disconnection is intended. 

About like this : 

Li The 

L. H. part is to be treated in the same manner in this andante 
i. e., both hands should be raised simultaneously. 
The manner of punctuating is similar here : 



Andante. Mozart. 

Ex. 46. 


The incomplete cadence is to be very slightly disconnected from 
what follows. Not more than ^ should be taken from the last 
eighth note at | . 

A different condition prevails when trie last notes of a phrase are 
staccato, as in this bourree from Bach : 

The first three notes in the second measure, being staccato, are 
played thus: ^'6 

Therefore it is evident 

that the method of punctuating applied to examples 44, 45, and 46 
would be ineffectual in Ex. 47. Rhythmic accent here plays an 
important part. An emphatic staccato is applied to the melodic 
tones on the fourth beat of every alternate measure whenever it is 
desirable to indicate the regular subdivision, thus : 

I ! 

The phrases (included within brackets) are by this simple method 
plainly revealed to the listener, and the bourree style is more strictly 

A summary of the preceding is here included. I. When the 
last tone of a phrase is of such duration as to forbid a pointed 
disconnection, the performer may subtract a fraction of its value 


(^ to y) as in examples 41, 43, 45. 2. Also, where the character 
of the music renders a staccato effect undesirable, the last tone of 
the phrase or group is left reluctantly, only a slight disconnection 
being discernible : Ex. 46. 3. In lively music, when the last note 
of a phrase is of brief duration, the staccato is-employed as a means 
of separating the rhythmic groups : Ex. 20, 21, 22, 23. 4. When 
disconnection is undesirable, an accent may be applied to the 
initial tone of certain motives or phrases. For instance, examples 
!7 33> 4- 5- If tne st yl e is staccato (thus excluding the ordinary 
punctuation, or rendering it inoperative), the subdivisions are 
marked by means of rhythmic accentuation. Ex. 29, 30, and 
such pieces as the " Pizzicati " by Delibes. Also the following 
more difficult works require similar treatment : " La Campanella," 
Paganini-Liszt ; " Le Tremolo," Gottschalk ; " On the Prairie," L. 
Schytte ; and the Staccato Etudes of Rubinstein, Max Vogrich, 
and Mme. Mazzucato-Young. 6. Phrases which commence after 
the first of the measure must not be so accented as to convey the 
impression that the initial note of a rhythmic group is likewise the 
initial note of a measure. In other words, an example like 48 must 
preserve its mensural and rhythmic qualities, because it is a 
peculiarity of the bourree that it begins upon the fourth beat and 
ends upon the third. Therefore care must be exercised not to 
create an impression like this : 

T? Af\ M Wr W~ r -\ Q-_ 17 /IQ 

Ex. 49. U^L- I 1 1 i Ui i i ^ J--J bee J^x. 48. 

The gavotte is still more liable to fall into this perverted men- 
sural arrangement, and that is why the author urges a very dis- 
criminating application of the general rule that beginnings of 
phrases and rhythmic groups "should be accented." Professional 
performers frequently fall into this error, and the author has heard 
several essays at Bach's G-minor gavotte in this style : 

Ex. 50. f^^lUJZ^tl Hth^t-Pf -t- ] Compare this with Ex. 41. 

- J 


PH RASI NG (Concluded) . 

Diminuendo. One of the means employed to indicate the close 
of a period is diminuendo. From several similar instances the fol- 
lowing is selected : 

Ex. 51. Andante from Sonata XV. Mozart. 

.li^^^^^j_.L_^^^j^j_ . ^^^^^^j_ ^^^^^^_ I- _ 

Under these circumstances no rallentando is made ; the harmonic 
cadence and the diminuendo being sufficient to mark the close. 
Also see the theme and variations by von Weber, op. 7, " Vien' 
qua, Dorina bella," and the B-flat Impromptu by Schubert. 

Rallentando. In lyric music the ritardando is frequently intro- 
duced at the close of a period, especially at the end of a move- 
ment. The adagio in Beethoven's op. 2, I ; the largo in op. 7 ; 
Chopin's " Berceuse " ; " Traumerei," by Schumann, and the 
Tempo di Menuetto, by Ph. Scharwenka, are familiar instances. 
Also, Grieg, op. 6, II, measures 15 and 16. 

The words a tempo usually follow the rail, or rit. as a means of 
indicating the entrance of a new (or repeated) period. 

Either diminuendo or rallentando is perfectly natural to the state 
of finality, or end accomplished, and frequently both are combined, 
as in Bendel's " Am Genfer See," op. 139, I, and the "Chanson 
sans paroles " by Tschaikowsky. However, it is not the present in- 
tention to formulate any theories concerning the application of 




these perdendosi effects, though under certain headings they will be 
duly considered. 

The Fermata. Where a period is completely closed, or where 
it has no immediate connection with what follows, the fermata may 
be introduced. Thus, in Saint-Saens' first G -minor Mazurka a 
pause may be applied to the last chord before the middle part in 
G -major : 

Op. 21, I. 

Ex. 52. 

Part II. 


, _ , i -I - p- 

Though there is some affinity between parts i and n, yet the two 
strains, (a) and (b), are so apparently dissimilar that by introducing 
a pause on the G-minor chord we give a broader termination to 
that part, and thus prepare the hearer for the change of style, mode, 
and rhythm in part n which follows. 

A more reposeful example of the use of pauses may be found 
in the little cradle song by Grieg, op. 38, I. The composer placed 
a fermata over the last note of almost every complete period. 
These tenuto signs indicate only a brief prolongation of the tones, 
excepting at the final close. 

In vocal and violin music the fermata is still more effective. By 
thus sustaining the last tone, and allowing it to die away, a certain 
effect of finish and repose is produced which is usually very satis- 
fying to the listener, even if not otherwise expressive. Indeed, the 
lack of this shows itself very unpleasantly in the singing of certain 
vocalists who, for want of sufficient breath, convey the impression 
that they are in a hurry to conclude the song, like an acrobat who 
has performed a dangerous feat and rejoices when the agony is 
over ! 

4 8 


Pauses are frequently included at the end of isolated arpeggio 
chord figures, either to represent the terminating point of a musical 
division or to give greater effect to an extended harmony. The 
Paganini Etude in E, transcribed by Schumann, op. 3, II, contains 
an instance. It occurs at the end of the passage immediately be- 
fore the final recurrence of the main theme. Referring to this 
arpeggio chord Mr. Sternberg, in afoot-note, says : " A pause, just 
long enough to allow the reverberation of the instrument to die 
out, is necessary here." The example is quoted in order to show 
the intended effect : 

Ex 53. 


The dominant chord continues to vibrate by means of open 
dampers, and when the sounds have nearly ceased the dampers are 
closed and the principal theme is resumed a tempo. The pause is 
particularly, essential here, because the passage (a) has no imme- 
diate connection (excepting this dominant chord) with the principal 
theme which follows at (b). Also, the fermata gives greater effect 
to the brilliant chord figure. 

While it is true that many instances similar to examples 52 and 
53 might be quoted, the student must not prolong the value of 
notes unless some very good reason presents itself for so .doing. 
For instance, no intermediate pauses should be introduced into 
Kirchner's Albumblatter, op. 7, excepting perhaps in the num- 
ber iv. An air of repose and finality may, however, be imparted 
to the last chord in nearly all cases. 


In the following extra selections the numbers refer to the sum- 
marized headings on pp. 44, 45, chapter v. 

I. Guirlandes, B. Godard, op. 107, XI ; Valse in A, Dvorak, op. 
54, I, (M). 2. Abends, J. Raff ; Am Meer, Schubert-Liszt. 3. 
Pierrette, C. Chaminade ; Papillon, Grieg, op. 43, I. 4. " If I 
were a Bird," Henselt, (D) ; Air de ballet, M. Moszkowski, op. 
36, V (especially last part in G -major) ; " Murmuring Breeze," 
Jensen-Niemann. 5. Air de ballet, Moszkowski ; The Chase, 
Rheinberger. 6. Mazurka, Moszkowski, op. 38, III ; Morning 
Serenade, Henselt, op. 39. 7. Chopin, op. 37, II. 

These are mostly of medium difficulty and can be had together 
in " Modern Musical Classics for the Piano/' 

The same volume may be utilized in illustrating the present 
chapter, thus : Serenade, Chaminade ; nearly all the periods are 
closed by means of diminuendo, and a few are marked "riten." and 
" rit." Ritardando, rather than ritenuto, is here indicated. Valse 
in A-major, Dvorak. Au Matin, B. Godard ; dim. and rail, ac- 
company all the cadences. The diminuendo is sometimes indicated 
by the dynamic symbol, ;^r== , which shows more plainly how 
far the diminish extends. 

Numerous instances of the fermata may be found. See " lunga 
pausa" in the " Callirhoe," by Chaminade. 


Nearly all composers have felt the necessity for avoiding the 
monotonous recurrence of regular rhythms by introducing uneven 
phrases, changes of measure, reversed accents, etc. The former 
will be treated here. 

As a general rule, phrases are even and contain two or four 
measures. But there are many instances of three-measure phrases. 
Subdivisions of this unequal character are less natural, and, there- 
fore, they should be clearly defined by the performer. The listener 
can not be supposed to entertain a preconceived idea of their exist- 
ence, nor to anticipate them, as he does equal phrases. 

Hungarian music affords the greatest number of uneven rhythms, 
so much so that to Saxon and Anglo-Saxon ears the effect is 
sometimes equivocal, if not disappointing. A few quotations are 
taken from Peters' " Czardas Album," No. 1487. The first three 
phrases in I are irregular. The fourth phrase, beginning^, is ex- 
tended to four measures. Then there is a phrase of three meas- 
ures and another of four. So far as the performer is concerned, his 
chief merit will consist in discovering these disproportionate groups ; 
their treatment presents no serious difficulty. The initial phrase is 

^=^ -^ 1 1. 

Ex. 54. 

The principal accent is to be reserved for the commencement of 
each phrase, at least until the rhythmical peculiarity has been im- 
pressed upon the listener. The three measures constituting the 
phrase must be considered as a unit as something thus far com- 
plete in itself. The unity must, therefore, be preserved by means 



of a continuous and connected performance. The third measure 
may be slightly retarded, or a brief pause made on the quarter- 
note, D. The second phrase is similar. So is the third, in length 
and in rhythm. Here a longer pause is to be observed. The four- 
measure phrase is easily managed because it leads naturally to the 
tonic cadence : 

Ex. 55. 

The second section, consisting of an uneven and an even phrase, 
is then repeated, and the lassu closes. 

One more quotation from this album is presented : 

Ex. 56. 




The slur in the second measure (not included in the original) is 
important,' because it helps to join the three measures together into 
a typical group. In similar manner the b-natural must be joined 
to its resolution, c, in the second phrase. All the groups in this 
adagio (lassu) are uneven. There are other disproportionate 
phrases among these czardas, and it would be well for the student 
to discover and mark these instances. 

The minuet in Mozart's great G-minor symphony contains some 
instructive examples of unequal phrases. The first section is 
presented : 

Ex. 57. 


A slight but animated staccato applied to the end of each phrase 
serves as punctuation of these three-measure subdivisions, though 
the accents also are important. 

This is followed by an eight-measure section, thus : 

-- ~ i.: 

The first phrase here is divided into two semi-phrases. The accent 
at 9 is, therefore, slightly more pronounced than it is at 10. This 
is verified by the full score, where we see that the flutes enter on 
the last of measure 8, thus making the phrases and semi-phrases 
even. A primary accent falls upon the first note at 1 1, which is 
an ending and a beginning. After thus marking the commence- 
ment of the last phrase the principal difficulty will disappear, 
because the cadence is so natural that it carries its own conclusion. 
The difference between the first section of six measures and the 
second section of eight is scarcely perceptible (when properly 
performed), so artistic and spontaneous is the conception. The even 
phrase runs into a free sequence descending (8, 9, 10), which 
counteracts the influence of the preceding three-measure groups. 
The last section also is modulatory, and this, add^d to the 
sequence, compels us to follow its wayward course by diverting 
our attention from the previous rhythmic formula. Being unable 
to forecast the result, we willingly leave that to the genius of the 
composer. Almost the entire minuet is composed of uneven 
phrases ; but the principal difficulty consists in outlining the even 

An even phrase may be made uneven by introducing an 
echo. Such instances are easily apprehended. In " Florian's 
Song," by Godard, an uneven phrase becomes even by means of 
an echo in the accompaniment. In these and all similar instances 
the echo belongs to the preceding, not to the following, phrase. 

In the Scherzo of Beethoven's last symphony there is a division 
in which all the phrases are irregular. To guard against the 



possibility of misinterpretation, these are marked at the beginning : 
" Ritmo di tre battute." With this understanding, the passage is 
easily managed, thus : 

Ex. 59. 

To facilitate further study of this subject the following list is 
.subjoined. " Czardas," by Jos. Low, (E) ; Minuet, from Haydn's 
" Oxford Symphony," (M) ; Turkish March, from " The Ruins of 
Athens," by Beethoven, (M), arranged by Rubinstein, (D). Also 
the Scherzo of Madame Clara Schumann, op. 15, IV, (MD). The 
sections contain five and seven measures. With exception of the 
initial period (which occurs several times) the other periods contain 
four-measure phrases. These instances should be sought out by 
the student, with the understanding that a correct performance of 
the music in which they occur is not possible until the melodic and 
rhythmic subdivisions are thoroughly comprehended. 


A musical period can not be regarded as finished without the aid 
of an authentic cadence, in some of its forms. In the analysis of 
a musical structure it therefore becomes necessary to understand 
the most important harmonic cadences. These necessarily include 
melodic cadence, since harmony without melody (even in common 
chord progressions) is not possible. 

Certain melodic and rhythmic conditions must be complied with 
before an air or theme can come to a satisfactory conclusion. We 
do not expect a cadence until four phrases (eight or sixteen meas- 
ures) have transpired, or until the theme has run its natural course. 
This is, of course, a primary statement of the case. 

As a general rule, the melody part is so influenced by its chord 
accompaniment that we will be obliged to take cognizance of the 
harmonic effect. To this end the principal final cadences are pre- 
sented in notation : 



10. 11. 


I. Simple forms of authentic cadence; a, b, and c are to be 
considered as harmonically identical. 2. Same, with the minor 
7th added. This tone, resolving down to the third of the tonic 
while the leading note ascends to the tonic, is a more positive form 
of this cadence. 3. Diminished 7th chord (II) used here in place 
of the dominant 7th. It belongs more particularly to the minor 
mode, since its tones are found in the harmonic minor scale ; but 
composers sometimes use it in a major key, as here. This is rather 
more serious than No. 2. Observe that the diminished chord at 3 
contains the two principal elements of an essential discord (4 and 
7 of any key), and that the minor 6th resolving down a half-step 
also assists in perfecting the close. 4. The dominant major gth is 
here used. It does not add any strength to the cadence, but may 
be classed among the final endings. 5. Similar to 3, but founded 
upon a pedal-note. 6. The dominant or dominant 7th chord pre- 
ceded by that of the subdominant (or any combination which has 
the effect of a subdominant harmony) constitutes a complete ca- 
dence, i. e. t it embraces every tone in a given scale. This close 
is, accordingly, still more conclusive and final. The classic com- 
posers, from Corelli to Mozart, used it in nearly all their final end- 
ings. 7. Relative minor of the subdominant substituted for the 
latter. 8. (a) A secondary 7th chord as subdominant here pre- 
cedes the dominant. A similar effect is produced at (b). g. A 
simple form of perfect cadence. The second inversion of the tonic 
chord (2) gives more smoothness to the progressions. Otherwise 
it is the same as 6. 10. This is a complete cadence with the ad- 
dition of a chromatic passing tone, suspension, and anticipation. 
II. Another kind of perfect cadence, using the passing diminished 
7th chord after the subdominant. 

Final Cadences in Minor. The fundamentals in all such har- 


monic formulas as I, 2, 4, 6, 8 are identical in the minor mode. 
For example, the foundation of a complete close in G-major or 

G -minor would be : Ex. 61. 


The key-tone here is completely established as that of G ; the 
mode would depend upon the prevailing tonality, or upon the fancy 
of the composer. In G-minor the close would be like this : 

Ex. 62. 


G-major the subdomi- 

nant and tonic would naturally be major chords, as they are minor 
here ; the dominant 7th chord is the same in both instances. 
The diminished 7th chord is a product of the minor scale : 

Ex. 63. 


This is a complete cadence. See No. 3 in Ex. 60. 
In a minor key the dominant Qth is naturally small : 

! J /TS 

Ex. 64. p-XL-p-p-Sig (g Compare this with 4. 

Y Y 

Incomplete (Half) Cadence. This reverses the order of an au- 


thentic close : Ex. 65. 

gE=E^_J ^- 




The dominant here follows tonic, and the effect is necessarily in- 
complete, something else must follow. It is the same in minor, 
and wherever the last chord is recognized as dominant. (The 
application of these will appear later.) 

Avoided Cadence. If, when a period is expected to close, the 
dominant 7th is followed by any other than the tonic chord, an 
avoided or deceptive cadence results, thus : 

Ex. 66. 

At the close of a section this would serve to prolong the period 
beyond its natural length, and therefore to postpone the point of 
repose. This is illustrated in the following excerpt from Mozart : 

Fantaisie in D. 
L 6 78 

Melodically the period closes at 8 on the tonic ; but Mozart 
intended to repeat the period, and therefore he did not write a 
regular close here : 

Ex. 68. 

8~ '~fi f f f I 

*#tt ? 

The B-tninor chord being substituted for that of D-major consti- 
tutes an avoided cadence, and from this we know that the end is 
not yet. In the repetition of this period there is a complete 
cadence ending on the tonic : 


Ex. 69. 

A few other forms of the avoided close are presented : 
In major. In minor. 

= *u 

Ex. TO. t3t=s!=:aite2Sr 

\ i 

After-cadence. After a movement has ended, the composer 
sometimes adds the subdominant harmony followed by that of the 
tonic, thus : 

Ex. 71. 

rit. ^ 


dei - ne mut - ter wacht. 



The after-cadence here forms a brief postlude to the song, and 
this example is especially selected because the esthetic application 
is similar in instrumental music. The after-cadence (known as the 
"Amen" in sacred vocal music) is mild and somewhat retrogres- 
sive in its tendency, and usually is played ad libitum. 

These various harmonic cadences have a determining effect upon 
period-construction and frequently influence the performance to a 


considerable extent. It is, therefore, absolutely essential for the 
performer to be familiar with these cadences in all major and minor 
keys, and to know the application and significance of the more im- 
portant harmonic closes. To this end it is recommended that exam- 
ples 60, 66, and 71 be transposed at the piano or organ into various 
major keys. Then the minor cadences should be treated similarly. 
The practical benefits which result from this theoretical work can 
scarcely be overrated. 

One of the most remarkable short examples, illustrative of this 
subject, is the Prelude IV, by Chopin. All the cadences are 
avoided, ^either melodically or harmonically, until the final ending 
in E-minor. We are led to expect a close at several points, but in 
each instance the hope is unfulfilled. The longing mood continues 
unsatisfied and unresolved. Finally, when the melody makes an 
apparent cadence on. E-minor (measure 21), the harmony avoids the 
tonic close and passes to C-major, thus : 

Ex.72, (a) 

Op. 28, IV. 

These avoided and deceptive cadences here not only prolong the 
periods, but serve a higher purpose in supplying the shadows to a 
picture in which never a glint of sunshine penetrates the gloom. 
The only authentic cadence occurs at the very close, thus : 

24 25 




It must be understood, with regard to the authentic, complete, 
and perfect cadences, that their most important application is at the 
end of a period or a form. It is solely by means of harmonic 
cadence, either expressed or implied, that periods are terminated 
or prolonged. In order to completely close a period the composer 
must employ the dominating harmony, or some principal discord, 
unless the melodic cadence is sufficiently determinate to suggest 
some of these. The incomplete, avoided, deceptive, and after- 
cadences each have their special application. 

In full harmony the cadence is more easily recognized than it is 
where the parts are few in number. In the latter instance a 
knowledge of chord representation is presupposed, since a single 
note may be intended to suggest a full chord. The Bach " Inven- 
tions " illustrate this : 

Ex. 73. 




A complete cadence in D is here plainly outlined. 
mony it would be thus represented : 


In full har- 

"Invention" IV is similar, and without exception the dominant 
will be found in the bass immediately before the final tonic. 


Chopin ended his G -sharp-minor " Prelude " without chords, 
thus : Ex. 74. 

The impelling force of dominant is such that this implied 
authentic cadence is entirely satisfactory ; the ear experiences no 
difficulty in comprehending the full harmonies of dominant and 
tonic. Hence the final closes in Bach's " Two-part Inventions " are 
satisfactory, because there is a melodic cadence above and a har- 
monic cadence below by means of the fundamentals. The latter 
are nearly always suggestive of subdominant, dominant, and tonic. 

The points where periods are closed or extended are usually 
determined by the harmony, and these cadences are of the highest 
importance in their influence upon punctuation, accent, and style.* 

The following works are recommended by way of elucidation : 
Bach's " Two-part Inventions," IV and X. Avoided and final 
cadences occur at the close of each ; the former were introduced 
with the sole object of prolonging the final close. 

In the first Scherzo, by Madame Clara Schumann, op. 10, 
(D), no cadence occurs until the i$th measure, and this is the 
beginning of the principal theme. Therefore, the introduction 
ends (according to the harmonic cadence) simultaneously with the 
beginning of the Scherzo proper. No completely closed period 
appears before measure 43. " Minuet," Beethoven, op. 31, III; 
the principal strain is repeated, and each time the cadence is 
incomplete thus forbidding a full close. The second period, also, 
is repeated, but here there are complete cadences. The first period 
of part ii terminates with an implied complete cadence. The 
second period ends with a perfect close. A. Hollander, " Concert 
March," op. 39, I, (D). Authentic and complete cadences ; also, 
after-cadence in part i. 

* Seventeen different species of harmonic cadence are illustrated in the author's 
" Analytical Harmony." 



Regular Period. Here it becomes necessary to inquire more 
specifically what constitutes a musical period, and how may it be 
distinguished ? 

In simple music the periods comprise eight (or sixteen) measures 
and are concluded by means of an authentic, complete, or perfect 
cadence. This agrees with the strict designation of period, as a 
close or a complete strain. The principal period of " The Watch- 
man's Song " is quoted : 

Ex. 75. 


Grieg. Op. 12, III. 


This is very simple and natural. The second period is equally 
regular and, like this, closes with a complete harmonic cadence. 

Where the period is repeated (either with or without alteration), 
composers frequently make the first cadence incomplete, so as to 
join the entire sixteen (thirty-two) measures into one complete 
strain. The op. 6, II, is in this style. The first ending, on the 
dominant, is harmonically incomplete, thus : 



Ex. 76. 


This leads to the repetition of this period in A-minor, at the end 
of which a more complete cadence occurs : 

Ex. 77. 

According to synthetic melody construction, the first eight 
measures comprise a short period of four phrases ; but, as the com- 
poser intends to repeat this, he leaves the first cadence (Ex. 76) 
somewhat indeterminate and joins it to the following. Then, after 
the strain has been repeated, the cadence is perfected and we hear 
a regular full close in A-minor. 

The performer must carry this idea into effect, and not give an 
impression of completeness to measure 8, but continue uninter- 
ruptedly to the more final close at 16. 

In the same composer's op. 12, VII, the cadences are similar. 
The second period (in G) does not end with its fourth phrase, but 
proceeds continuously to the repetition. Even this repeated period 
of 1 6 measures does not terminate harmonically, but by means of 
an incomplete ("half") cadence it leads naturally to the return of 
the main theme in tonic minor (a) : 

6 4 


Ex. 78. 

Measures 23, 24, form the last phrase of the repeated period 
beginning in G. The incomplete cadence here serves to unite the 
two strains. Therefore the periods at 16, 24, 40, and 48 should 
be implied, but not expressed. Hence an important distinction is 
to be made between melodic and harmonic cadence, for the latter 
colors and otherwise modifies the former to a considerable degree. 

The " Air de Ballet " by M. Moszkowski presents several interest- 
ing features for analysis. The first period of eight measures has no 
harmonic cadence, and therefore continues uninterruptedly beyond 
this point. But the second period begins unmistakably upon the 
9th measure, and this must be indicated by the performer without 
regard to the completeness or incompleteness of what precedes. 
These seeming contradictions between theory and practice con- 
tinually present themselves to students who attempt to analyze 
musical construction. But by understanding certain principles and 
observing certain distinctions the problem will finally be solved. 
Theoretically, the musical period may be said to embrace the melodic 
embodiment of a theme an entire strain. It is the natural develop- 
ment of a motive to some satisfactory, if not logical, conclusion, as,, 

Ex. 79. 

for example, in the " Air de Ballet " under notice. The motive is. 
continued in melodic development during eight measures, and then 
a second period begins. Therefore the first period (without regard; 


to mensural proportion or harmonic cadence) forms an independent 
strain according to outline analysis. The first eight measures com- 
prise all that the composer chose to say upon this phase of the 
subject. But if we examine this period harmonically, it will appear 
that no cadence is included in the last phrase ; and therefore there 
is no close, no point of repose, at the end of this melodic division 
of the work. Hence the conclusion is quite logical that the 
music proceeds without interruption (a tempo), except that the 
second period must be duly marked in order to indicate its advent. 
This second period (a development of the rhythmic feature of the 
motive) ends with a complete cadence on the i6th measure : 

Ex. 80. < 

Here the composer says "rit," and the accompaniment makes a 
natural return to the first strain marked a tempo. According to 
the strict signification of period, as it is to be applied by the 
performer, this cadence (15 and 16) presents the first instance of 
complete period. This view is confirmed by the composer's own 
explicit directions. .. The two strains following (similar to the first 
sixteen measures) have no harmonic cadences, and consequently 
they are incomplete periods. The second of these is joined 
without interruption to the second theme, in G-major, which must 
be duly marked at its commencement. And every independent 
strain, whether separated from or connected with its antecedent, 
must be plainly indicated by means of accent, or some noticeable 
alteration of the movement usually a tempo. 

The second part continues (with only a brief cadence at 40) as 
far as measure 53, where there is a pause after the pedal-note 
discord in order to separate the preceding from the ad lib. cadenza. 



Then the first part recurs as before, and is joined to a brilliant 
termination in G-major. The first period of this is regular in 
construction, though continuous and uninterrupted save by accent. 
The repetition of this, beginning an octave higher, modulates freely 
and partakes of the character of passage-work. There is no 
cadence and no period until we come to this : 


Ex. 81. 

Then there is a continuous strain of twenty-four measures, ending 
on G. The remaining fifteen measures form a stretto, without 
intervening cadence. 

Other illustrations, tending toward a more thorough understand- 
ing of this misapprehended subject, will appear farther on. There- 
fore may we proceed with the consideration of 

Curtailed Period. This is not of frequent occurrence in 
modern music, excepting in strettos and other final passages. 
There are, however, a number of instances, and some of these are 
rather difficult of management. Where the curtailed period closes 
a movement, it presents no obstacle, because irregularity of con- 
struction is here in order. The final tutti in the first allegro of 
Hummel's A-minor Concerto, op. 85, is quoted: 

Ex. 82. _^ 

^^^ fe. ^h^^ ^ '^^^ W. ^^^^^^i^^^^J ^^l^^^^^l 

i ~ h ^~ ! i n i 

-I h 



6 7 

* j- = -H - 

_ 1 

The continued repeating of tonic and dominant destroys the 
impression of regular construction by phrases ; and, besides, this 
is the final stretto. The effect is perfectly satisfactory because the 
irregularity (seven measures) is not noticeable. 

A curtailed period occurring intermediately is not so easily 
managed : 

Ex. 83. Largo. Beethoven. Op. 2, II. 

I ;=!; 1 1 I 

^^N-| *-m 

* if > V 

The peculiar turn of the perfect cadence (measures 6 and 7) 
is so impelling as to produce an effect of completeness, though 



this period lacks one measure of the full number. (See first eight 
measures of this largo.) The composer's directions aid materially 
in the interpretation ; but, in addition to these, the author recom- 
mends that the usual method of expression by phrases be ignored 
here. The transitional sequence in measure 3 (not found in the 
first period) tends to disturb the former equal rhythms and seems 
to belong more to the second than to the fourth measure. The 
aim should be to unite the remainder into a complete section. A 
gradual ritardando at the close will contribute to this effect. 

A more peculiar instance is here cited from Grieg's op. 3, III : 

^>s ^^=pftgzr=g;gin 

b \t v L I I r 


r*N^^ * \A -* -*-* HI -+~i 

The movement here being much less slow than it is in example 83,. 
we are more inclined to notice the irregularity of this curtailed 
period. The perfect cadence at the end is, of course, very posi- 
tive and conclusive ; still, the termination of the period is inclined 
to sound premature and somewhat disappointing unless the 
nuances are managed judiciously. Two methods of interpretation 
are available: (i) Consider the third, fourth, fifth, and sixth 
measures as a continued thesis, to be played without regard to 
phrase divisions ; then, after a brief pause, sound the last four 
chords more slowly and with very decided emphasis. (2) Make an 
uneven phrase of measures 3, 4, and 5 ; pause at the end of this, 
and then perform the last two measures rallentando as an even 


phrase. Also join the last eighth to the first of the following 
measure, thus : 

Ex. 85. molto rit. 

poco rit. A A A A 

In either case the ritard and forzando in the last measure are to 
be included. The former method, in which the last measure rep- 
resents a final phrase, is more capricious and, therefore, suited to 
the fugative character of this little tone -picture. The latter plan is 
more logical and represents a rather serious mood. The author 
inclines to this interpretation, but another choice is freely left to 
individual fancy. This curtailed period occurs in the first part and 
again at the close. It should be compared with the eight-measure 
period immediately preceding. 


Regular periods : Grieg, op. 6, I and II ; op. 12, IV the 
principal theme ; Ph. Scharwenka, " Tempo di Menuetto," op. 55 ; 
Raff, "Fabliau." 

Curtailed periods : Grieg, op. 6, III two instances ; Beethoven, 
op. 7, "Largo"; op. 13, Adagio. There are several periods of 
six and seven measures, one of these being an Eingang. Dvorak, 
" Valse," op. 54, I ; the Eingang contains twelve (in place of sixteen) 

The selections from Raff and Grieg are rather easy ; the others 
are more difficult. 



United Period. In concerted music one part frequently begins 
exactly as another part ends. Since conclusion and commence- 
ment are thus simultaneous, the usual result is a slight contraction 
as regards the mensural proportion of the two equal periods, one 
measure being counted twice in the enumeration. To all such in- 
stances the author applies the term " united period." This presup- 
poses that the new period does not wait until after the previous 
one has ended, but enters simultaneously with the last note of the 
latter. An example is quoted from the rondo of a well-known 
piano concerto : 

Ex. 86. 

Hummel. Op. 85. 

-_-_- _ 

'- \ ^- E^ fa '' U " |h ~; ~^~j rb :: ^'~ ~ 

Only the cadence of the solo part is extracted, and as this ends 



on the tonic the orchestra enters simultaneously with a tutti 
passage, which forms a new period. The united period takes place 
exactly at this point, indicated by the symbol used in the author's 
" Complete Musical Analysis," P. U. 

The new period in all such instances is to be distinctly marked, 
because we are not inclined to anticipate it until after the previous 
period has been brought to a complete close. When the conclud- 
ing passage has been retarded in its cadence the regular move- 
ment should, as a rule, be promptly resumed as the united period 

Several similar examples occur in this concerto and in all music 
of this class. 

When the two periods are executed by a single performer, the 
difficulties are increased, both in the analysis and in the interpre- 
tation. The piano part alone does not show so plainly as does the 
full score of a concerto the dual nature of a united period ; nor is 
the responsibility shared with other performers when one is playing 
a solo unaccompanied. 

A simple illustration of this is here quoted from a sonata by 
Reinecke : 

Ex. 87. 


P. U. 

A period closes on the tonic at (a), and here a new strain begins 
at the same time on the 5th of the key, (b). Since all such in- 
stances presuppose or suggest different instruments, it would be 
well to imagine this in simple score, thus : 


HORN. (Sounding as written.} 

ding as writt 

__ c L zzbr i J^bbr L|- 

Observe, as the first period ends, that all the string instruments 
rest, with exception of the 'cello, which continues as base to the 
horn solo. The distinct accent mark, as indicated here, is essential 
in all similar instances. Furthermore, it is to be stated that the 
new period (b) is not a mere continuation of an unfinished one, but 
an independent passage or strain commencing simultaneously with 
the cadence of a completed antecedent period. 

Extended Period. This presents one of the greatest obstacles 
to outline phrasing, because the extension of a period usually dis- 
turbs the symmetrical order of period-groups. The prolonged 
period is, however, an essential feature of all but the simplest 
music. It prevents monotony, increases the interest, and prolongs 
the intervals of action and repose. It should be understood that 
the extension begins from that point where the regular period 
would (otherwise) naturally end. Transition, avoided cadence, 
repetition, and passage are the usual means employed in enlarging 
a period. A simple illustration is taken from Grieg's op. 3, V. 
The first complete period embraces eighteen measures, constructed 
in this manner : The first eight measures are left without cadence, 
in order to lead more continuously to the repetition, which con- 
tains ten measures. The extension is produced by means of an 



.avoided cadence to D-minor (in place of F-major) at 16. The last 
section is quoted : 

15 16 

Ex. 89. < 

a-Q 1 

The last phrase (extension) is a melodic repetition of 15 and 1 6, 
and after the avoided cadence this repetition becomes necessary in 
order to make the close of the complete period perfectly satisfac- 
tory. There is a brief pause here at this moment of repose. A 
similar example occurs in the second period of Tschaikowsky's 
Cliant Sans Paroles. The last six measures of this period are 
quoted, showing the method of extension : 

Ex. 90. 

a tempo. 
11 1-2 


t=l EPftadyndtair 

The usual length of periods in this song without words is eight 
measures, but there is no attempt at cadence in the 7th, 8th, or Qth 
measures of this strain. By repeating the fourth phrase at 9 and 10, 
and adding a measure (i i) for the close, there results an extended 
period (P. E.) of twelve measures, ending on the tonic at a tempo. 
Here the principal theme is resumed, being united with the pre- 
ceding at 12. The third period, beginning in canonic style, pre- 
sents a similar instance. The cadence is a deceptive one, but that 
does not prevent the tenor-theme from being distinctly indicated 
as it enters. 

With regard to the performance, it should be understood that 
all these extended periods are to be integrated as a complete 


strain, and not isolated as though the added measure formed a 
Codetta or Eingang. In the examples quoted it has been demon- 
strated that no cadence (and thus no period) occurs at the usual 
points of repose where we would naturally expect to find it, 
according to mensural proportion or regular period-formation. 
And since the composer's evident intention was to prolong the 
period, the performer must be governed by this purpose, and not 
seek to create an impression that the period has closed in the 
midst of a continuous, connected strain. When the nature of the 
extended period is comprehended, it will be comparatively easy to 
catenate the entire strain or passage, either by means of ritardando 
or crescendo, and the avoiding of primary accents such as are used 
to indicate the beginning of certain periods. In such instances as 
the tenor-theme in Tschaikowsky's Chant Sans Paroles the performer 
is to apply a more marked accent than is required to indicate 
mere phrase-divisions : 

This would also apply to the " Air de Ballet," by Moszkowski, 
especially to the beginning of the stretto, thus : 

Ex. 92. 

because this is the commencement of an independent passage, and 
the prevailing style is already very animating before the stretto is 

The points of repose in the Chanson Sans Paroles are three in 
number, to wit : measure 16, slight rail, and dim.] 35, rather more 
reposeful ; the final close, morendo. 

A continuous legato (where the style admits it) may likewise be 



employed as a means of connecting the extended period and there 
joining it to the cadence. 

Before citing another instance it will be well to consider the 
difference between an extended period of, say, 12 measures and a 
regular period of 8 measures with the addition of a four-measure 
codetta. An example of the latter is quoted from Mr. Sternberg's 
" Night Song," beginning with the second period, in B-flat. The 
phrases begin upon the third beat and end upon the second : 

Ex. 93 Sternberg. Op. 56, VI. 


inn Quasi adagio. 

* . . --^^ 

-*-! 1 ^5^*^^ - 

-*-*-*-: 555= 

a tempo. 

8 b. 

H* f * * . ? 5P5RfTT ?~~ 

poco ritardando, 


Eg. ,^_ . s 9 -^_ , K 10 r~3 

A regular period begins here at (a) and ends on the dominant at 
(b). Upon the last of this measure, (8) after the eighth rest, there 

7 6 


begins a short chromatic Eingang in form of a codetta, for the pur- 
pose of leading naturally to the return of the principal theme in 
E-flat, cantabile. The codetta of four measures (c to d) is sepa- 
rated from the previous completed period : (i) By the rest at (^) ; 
(2) by the difference in compass ; (3) by the episodial nature of the 
chromatic passage, which is in contrast to the preceding. The 
regular period in B-flat is extremely gentle, whereas the Eingang 
is darkly colored and somewhat agitated as it forcibly rises to the 
harmonic climax. The 8th rest between the regular period and 
the Eingang (placed there by the composer) is therefore a necessary 
punctuating mark, and the performer must feel that two different 
moods are here expressed. 

Now compare these first twelve measures (regular period, 8, 
and codetta, 4) with the continuous, extended period from Tschai- 
kowsky's Chanson Sans Paroles, previously quoted. The extended 
period begins : 


Ex. 94. 


and ends : 

Ex. 95. 

In each instance the music tells its own story, after the design has 
been analyzed. 

This charming nocturne by Sternberg contains an example of 
extended period also. It commences with the repetition of the 
main theme after the Eingang, quoted in Ex. 93. The cadence does 
not occur until the 25th measure from the beginning of the period, 
cantabile. This finely conceived prolongation is in the style of an 
endless melody, and the performer need not seek to measure it out 
by arbitrary phrase rules, as is too frequently done. A continued 



thesis, such as this, is above and beyond the reach of arbitrary 
formulas. If we follow the spirit of the music, the responsibility 
will rest upon the composer, and surely he is the better judge. In 
the present instance, though the period as a whole is uneven (25 
measures), the harmonic sequence is so conceived as to lead most 
naturally to the full cadence on E-flat at a tempo. All primary 
accents and pointed disconnections, as punctuating marks, are there- 
fore to be omitted during the prolongation of this period. 

A similar instance may be found in Rubinstein's op. 3, I. It 
occurs in the last repetition of the main theme, after the second 
intermezzo. The complete period contains 25 measures, an exten- 
sion of 9 measures, though the entire 25 are to be treated as a 
complete whole, and not isolated before the terminal cadence. This 
should be compared with the repetition of the principal theme 
immediately following the first intermezzo. 

The extended period from Rubinstein is not in the style of a 
continued thesis, and therefore is more susceptible to rhythmic 
accentuation than is the extended period in Sternberg's " Night 
Song." In the melody from Rubinstein the last section of five 
measures is repeated by means of a deceptive cadence, which makes 
the termination more emphatic. 

These instances must not be confused with those in which a 
repeated period is prefaced with four measures as a diversion. An 
extract is made from Beethoven : 

Ex. 96. 


Bagatelle. Op. 33, VI. 

^_i_Jj_! ^ 


Apparently the first four measures are the beginning of a second 
period ; but at the end of the section (12) the initial period appears 


and is exactly the same as in the beginning. The intermediate 
matter, (a) to (b), is really a relief to, or digression from, the main 
theme, which recurs several times. 

And even if we attempt to perform the twelve measures as a 
whole, the listener will immediately recognize the principal theme 
at (b) and associate it with the first eight measures, which are 

The Boccherini Minuet in A presents a similar instance ; so does 
the minuet-rondo in Beethoven's op. 49, II. 

With regard to other examples of united period, the student 
would better seek them in the piano concertos. Most of these are 
necessarily difficult, but the one in G by Hummel, op. 73, and 
Beethoven's first, in C, are comparatively easy. After observing a 
number of instances (such as Ex. 86) the student will be enabled 
to discover other examples in solo works. 

Almost every modern work, not in the common dance form, 
contains an example of extended period. See " Dance of Elves," 
Grieg, op. 12, IV; " National Song," op. 12, VIII (period extended 
by means of avoided cadence); Impromptu, Schubert (in A-flat\ 
op. 142, II, second period ; Sonata in C, XV, Mozart, first and 
second subjects. 


An important feature of this system consists in applying the 
peculiarities of minor details to practical performance. The 
author first directed attention to the minutiae of musical composi- 
tion and employed a series of symbols for use in theoretical 
analysis. These devices and details include every peculiarity that 
may be observed in the design and construction of music. The 
most important of these will be illustrated in what follows. 

i. Prelude. Introduction. For purposes of auricular analysis 
the author distinguishes between prelude and introduction ; but the 
main requirement here is to know whether a principal theme begins 
at once or is prefaced with introductory matter. The usual 
purpose of a prelude is to indicate certain features of the music 
before the theme commences. These are : Measure, movement, 
mode, rhythm, or style of accompaniment. If the object is to 
establish the measure, movement, and accompanying rhythm, then 
we must so impress these characteristic features upon the listener 
as to leave no room for doubt. The first two measures of 
F. Hiller's " La Ronde de Nuit" afford a simple illustration : 

F. Killer. Op. 146. 

Con moto. 

Ex. 97. 

This brief prelude determines the measure, movement, and mode, 
and should be performed with strict mensural accentuation, though 
lightly. The composer's directions are suggestive : un poco marcato, 
sempre egualmente. 




Moreover, the figure below constitutes a ground-base which 
continues throughout the rondo. It is, therefore, of considerable 

But where the prelude indicates merely the measure and 
movement, the strict mensural accentuation is to be relaxed or 
modified after the principal theme enters. (See examples 6 and 7, 
chapter n, together with the preceding and following remarks.) 

Another fact must be determined in reference to the prelude or 
introduction : Whether the preliminary matter is separate from, or 
leads naturally into, the main theme. 

The prelude to a Hunting Song by J. A. Jeffery, op. 7, is an 
instance of the former kind. At the end of the fanfare there is a 
pause on the essential 7th chord, because the prelude is quite inde- 
pendent of the hunting song proper. 

The introduction to " La Fileuse," by Raff, and " Au Matin," 
by B. Godard, present similar instances. On the other hand, the 
prelude to Chopin's Mazurka, op. 3, III, runs naturally into the 
chief melody, measure 9. The following are similar in this respect : 
Fantaisie, op. 16, I, Mendelssohn; Chant Polonaise, Chopin-Liszt; 
Spinning Song, Otto Hackh, op. 50. 

2. Antiphonal Groups. The term " antiphonal " is here ap- 
plied to all responsive phrases and semi-phrases, but not to every 
instance of arsis and thesis. The second group of the antiphone 
is usually of a negative character (like antithesis), and seems to 
issue from another instrument. Thus, from Mozart : 

Ex. 98. 


-i h-i -^* 

Each phrase lies in a different register one is loud, the other is 
soft ; the first is bright and positive, the second is rather serious 



and regretful. The style of performance is therefore influenced by 
these conditions, and may be indicated thus : 

Ex. 99. 

Though the two phrases would be scored for different groups of 
instruments in an orchestral arrangement, it is unnecessary here to 
particularize further. 

In the next example the semi-phrases are antiphonal, and require 
strong contrasts in quality as well as quantity of tone : 

Ex. 100. 

P. E. Bach. 

The opening salvo, rather bold and ponderous in style, is 
responded to by a rapid arpeggio figure (b), legato and piano. The 
former should, therefore, be well separated from the latter, and a 
marked difference in tone-quality is to be observed ; though the 
melodic sequence below is not to be neglected. 

Ex. 101. 

A light harp effect would be appropriate to the responses (b) and 



The following antiphonal section from Haydn shows still more 
plainly the contrasting phases of a musical period : 

Ex. 102. 


Minuet from "Oxford Symphony." 

TZzr -- 

The diversity of style between the opening phrase (a) and its 
response (b) is sufficiently marked, even without the orchestral 
indications. In such instances the task of the pianist is a simple 
one. Good taste will suggest that the scale passage be played 
smoothly and with very little accent. 

3. Echo. The first illustrations of this natural phenomenon 
will consist of a phrase or semi-phrase repeated an octave higher, 
and separated from the original figure which is echoed. The echo 
suggests an additional voice or instrument of a lighter but other- 
wise similar quality. 

The peculiar effect of reverberation (and of greater or less 
distance) must, usually, be imparted to the echo, especially where it 
is supposed to represent natural conditions. 

An example is quoted from Beethoven : 

Beethoven. Op. 49, I. 

Ex. 103. 

_ . i^g jf ^_ _i 

^^^^^^^^ i ~n i 

jo (Echo.) 

The echo here is to be played less distinctly than the original 
phrase, not alone on account of the repetition, but because the echo 


is naturally veiled. When the repeated figure does not follow its 
antecedent so closely we may suppose that a greater distance inter- 
venes between the original sounds and the point of reverberation. 
This would render the echo still less distinct. The " Souvenir de 
Suisse " illustrates this point: 

J. Raff. 

4-f *** i * 

Ex. 104. r -** 

The f and // were included by the composer, who undoubtedly 
had in mind a very distant and rather faint reverberation. 

A charmingly realistic effect is produced in Meyerbeer's " The Star 
of the North," where Elizabeth and the troops depart in boats from 
the Gulf of Finland. It is the scene in which the favorite prayer 
and barcarolle occur : 

" Let echoes tell 
Our sad farewell." 

Echo must not be confused with canonic imitation, which will be 
explained later. 

4. Sequence. It is scarcely possible to conceive of a compo- 
sition in which sequence does not occur, either melodically or har- 
monically. Melodic sequence consists in repeating upon different 
degrees of the scale a motive or group in continuation of a certain 
theme, as here : 


Ex. 105. 

i rrv) "" 


Design. Sequence. 

The group (a) is sequenced at (b), (c), and (d). The continua- 
tion of a sequence, however extended it may be, is a sign of con- 
nection or unity, since this natural order of following carries with 
it the same thought or sentiment. Another consideration which 


applies to practical performance is that the sequence is delivered 
by the same voice or instrument which announced the design, and 
therefore this affinity must be maintained. 

The opposite of this is true in the antiphonal style and in the 
echo, as we have seen. Where the sequence transcends the limits 
of a certain compass the composer may employ another instrument 
in the continuation. But in such an emergency he would choose 
instruments of the same class to deliver the design and the 
sequence. The statement may therefore be reaffirmed that, in 
respect of tone-quality, no marked distinction is to be made during 
the continuance of a melodic sequence. 

In the majority of instances it will be found that sequacious 
passages increase as they ascend, and decrease as they descend. 
This is particularly true in the music of nature, and the principle 
may be applied tentatively by young performers in their first 
attempts at expression. But in such a scene as, for instance, the 
Turkish march in Beethoven's " Ruins of Athens," the dynamic 
conditions are influenced by supposed distances, and these imag- 
inary actualities must take precedence over all arbitrary formulas. 
As the soldiery approach a given point the sounds gradually 
increase in volume ; then the music diminishes during several pages, 
until it is seemingly lost in distance. This, of course, represents 
the passing of the military out of view. Material considerations 
and physical conditions thus control the increase and diminish of 
tone, without regard to ascending or descending sounds. The 
remarkable effects which Rubinstein produced in this Beethoven 
work revealed the almost unlimited extent to which dynamic con- 
trasts can be carried in artistic performance. 

The same observations apply to harmonic sequence, though a 
distinction is to be made between the free and the strict species. 
These are fully illustrated in " Analytical Harmony " (chapter 
LXIII) and need not be enlarged upon here. 

5. Anticipation. This is not to be understood in its harmonic 
sense, but refers to a preparatory group of notes which serves to 
introduce a principal or secondary theme. For example, in the 
Cabaletta by Lack, where the theme recurs the last time there are 
several measures of anticipatory matter introduced. The first three 


notes of the theme (quoted in Ex. 29) are selected as motive for 
the anticipation (a), and this is repeated in sequence until it leads 
naturally to the principal melody at (b) : 

Ex. 106. 

*_. _ _ __ 

I i 1 L* 

All such examples the author designates as anticipations, because 
they enable us to forecast, as well as to expect, the ensuing strain. 
The anticipation is of an impatient character, and nearly always to 
be played crescendo or accelerando. In the last example cresc. 
implies a slight increase in movement as well as in tone. 

Other simple illustrations may be found in the Polonaise by 
Reinecke, op. 47, III ; in the Rondo to Beethoven's op. 49, I ; and 
a more difficult example in the D-flat Waltz by Chopin. The four 
introductory measures were evidently intended by the composer to 
anticipate the principal theme, (b) : 

On. 64. I. 

As the movement of this waltz is extremely fast it would be well 
to begin the anticipation allegro and accelerate the speed to allegro 



After the second subject this anticipation recurs, prefaced with a 
long trill. 

6. Canonic Imitation. Free. Canonic imitation may, as is 
well known, take place upon any diatonic interval. A brief example 
in the 7th is presented : 

Ex. 108. 


r-tr r^^r 

The second voice here imitates the first at the interval of a 7th. 
This is free imitation because the large 3d at (a) is answered by a 
small 3d at (b), and so on. In like manner (c)-is imitated at (d), 
and (e) is imitated at (f ). The theme in the answering voice-part 
bears such a close resemblance to the upper melody that the former 
is to be made almost as prominent as the latter. To be precise, 
the imitation below is the main theme slightly altered in order to 
preserve the same tonality.* Imitations in the 2d, 3d, and 6th 
also are free. 

7. Strict Imitation. Imitations in the unison and octave are 
necessarily strict, every interval in the answering part being theo- 
retically identical, thus : 

Ex. 109. 



-*=. *r*: 

* It is not here necessary to explain all kinds of imitation ; that is done in such a work 
as Cherubim's "Counterpoint and Fugue." But the principles should be well under- 


The two voice-parts are equal in melodic importance, and there- 
fore they are to be performed accordingly. And, moreover, the 
interest is to be well sustained in the canonic style, for while one 
part makes its cadence, the other part (or parts) will be in the 
midst of a phrase or section. 

8. Contrary Imitation. In this species the direction of the 
theme is reversed by the answering voice ; ascending tones being 
responded to by descending ones, and vice versa : 


I ^** | I e ^' 

W--F f- ! f 

The risposta (b) may be compared to a dissenting opinion, every 
interval of the proposta (a) being in reverse order at (b). 

As a general rule, this species will require more special accents 
than are necessary in the other styles of imitation, for this reason : 
that a disputatious argument is more animated and emphatic than a 
mere conversation in which no contention is manifest. Strict imita- 
tion is affirmative ; contrary imitation is negative. They belong to 
opposite states of mental activity, and are worthy of more consid- 
eration than they have thus far received from performers. Wagner 
employed this device in his music-dramas, and special significance 
attaches to these instances whenever a leading motive appears 
inversely. The " Mime " and the Compact motives in Siegfried 'are 

Partial or interrupted imitation does not require specially differ- 
ent treatment, though numerous instances will present themselves 
as we progress. 

9. Rhythmic Imitation. This is, of course, independent of 
melodic considerations, and refers to the actual value or rhythmic 
arrangement of notes in a given motive. In musical development 
rhythmic imitation is frequently an important element, and con- 
sequently it must receive attention from the performer. The 
allegretto from Beethoven's /th symphony is an instance. Dur- 



ing the A-major portion, where the minor theme has apparently 
been superseded, the bases persistently maintain this rhythm, 

I I I lj thereby preserving a characteristic feature 

of the original monotone motive : 

In orchestral music it is customary to assign these short motive- 
rhythms to the instruments of percussion, as thus, in the Choral 
symphony : 

Ex. 111. 


It is not always necessary to employ a monotone in rhythmic 
imitation. Where the composer desires to maintain a central idea 
in the midst of a varied melodic theme he may use the rhythm of 
the chief motive with good effect. In Th. Kirchner's 8th Album 
leaf the motive is this : 

Ex. 112. 

At the close the base maintains this rhythm, as at (a) and (b) : 

Ex. 113. 





Op. 7, VIII. 



. _ 

Especially at (b) do the rhythmic imitations preserve the unity of 
design against the contrasting theme above. Where the rhythmic 


8 9 

design of a motive is thus characteristic, it is frequently developed 
more prominently than is the melodic feature of the motive. 
Compare the first and second periods of Moszkowski's Air de 
Ballet ; also the A-flat Impromptu by Schubert, and the second 
Album leaf by Kirchner. 

10. Parenthesis. This belongs to the graces and ornaments of 
music, and calls for the exercise of refined taste. It is to be under- 
stood literally as something parenthetical ; usually a group of 
unaccented notes occurring between two measured melody notes : 

Ex. 114. 

Chopin. Op. 55, I. 


The parenthetic group might here be omitted without detriment 
to the principal melody. The parenthesis may therefore be con- 
sidered as somewhat adventitious, though it adds to the charm of 
the music and says something (entre nous, as it were) which is not 
told by the plain measured notes of the melody. 

In Ex. 114 the parenthesis is not to be understood as a mere 
conclusion to the trill, to be executed a tempo. But the trill is to 
continue after the so-called Neapolitan 6th has been sounded on 
the 4th beat. Then, in order to avoid the effect of precipitancy, 
the parenthesis is to be interpolated softly, yet distinctly, and with 
moderate speed. The movement must therefore be slightly re- 
tarded here, and in the following perfect cadence. 

The next quotation, from Beethoven's ist piano concerto, is 
similar in design, though written mensurally : 

Ex. 115. 


9 o 


The original arrangement appears at i. After the third tutti the 
theme is embellished and the parenthesis shown at 2 is introduced. 
These parentheses are unaccented, and they are usually to be 
understood as a species of punctuation. 

The works of Chopin contain more examples of this sotto voce 
embellishment than do those of any other composer. See his 
Nocturnes, IV, V, X, and XI, especially the parenthetic groups 
in small notes. 

II. Counter-theme. This is derived from fugal construction 
where the continuation of subject or response usually becomes 
counter-subject, and serves primarily as counterpoint to the main 
theme. When thus employed the counter-subject is so conceived 
that it will go with the subject either above or below, forming what 
is known technically as " double-counterpoint." Such examples 
afford the best illustration of counter-theme, and therefore is the 
following excerpt taken from a fugue : 

Ex. 116. 

< a >- C. S. 

Cat Fugue. ' ' D. Scarlatti. 

At (a) the subject is below and the C. S. above. At (b) the two 
themes appear inversely in regard to this order. The C. S. is so 
different from the subject that no effort is required to distinguish 


each part. Every time the subject appears it is accompanied by 
this syncopated C. S. 

Counter-theme is next in importance to the theme. Their rela- 
tive degrees of tone-quantity when combined may be thus ex- 
pressed : Theme, mp., counter-theme, /. Or, theme, / C. S. mf. 
When the C. S. is below, and especially where it is not so charac- 
teristic as is the main theme, then the former may be played with 
equal force. This is true of Ex. 116 (b). (The misapprehension 
which many students have of the character of counter-subject is 
owing chiefly to the false doctrine which has been disseminated by 
so-called educational writers. Thus, in a program book of Bach 
illustrations we read : " Let this counterpoint (counter-theme) be 
soft and the melody loud." Some excuse should perhaps be made 
for the penny-a-liner who put forth this absurd notion ; but the 
results of such doctrine are almost as pernicious as though it were 
promulgated by a musical authority.) 

The C. S. is frequently developed into a regular theme, as in 
several of Bach's fugues and in the close of Handel's F-sliarp 
minor fugue, Clavier Suites, Peters' edition, 1058. 

All examples similar to the following tenor melody are to be 
considered as counter-themes and treated accordingly : 

Ex. 117. 


\ ^~ 

The C. S. here is not an entirely independent theme, but was de- 
signed to harmonize with the soprano part above. Still, the tenor 
part possesses sufficient melodic character to demand recognition, 
and should be played slightly marcato. 

Certain melodic progressions in the base come under this general 

9 2 


heading, though in strict designation they are not counter-subjects. 

Observe the following from Chopin : 

Op. 55, I 

The melodic design below is plainly indicated *and this requires 
a light accentuation, especially as there is nothing else against the 
cantabile theme above, quoted in Ex. 114. The ad libitum parts 
(chords in the middle) are considerably subdued, but the real-base 
part ought to leave an impression like this : 



The next selection shows more plainly the distinction between 
principal theme and counter-theme : 

Ex. 119. Sonata. Violin and Piano. Rubinstein. Op. 13. 

C. 8. 

' ^zzr.: Z= S ? 


sgi =M 

i mg^tami ^^^ I M ' 

b^ ' 4 ^~~ J** ^- 




The piano part consists of the leading motive in development, 
while the violin has a rather serious counter-theme. This is to be 
one degree softer than the theme of the piano part, which is about 
mp. Both parts are marked /, as is customary in scores, and the 
performers must therefore analyze the entire work in order fairly 
to represent the composer's idea. This is why we hear so many 
unsatisfactory performances of chamber music. The distinctions 
which ought to be made between subject and C. S., canonic imita- 
tion, anticipation, and ad libitum parts are either overlooked, or 
not properly apprehended. 

Music of the harpsichord epoch contains innumerable examples 
of C. S., but the style is mostly polyphonic, and that will be con- 
sidered hereinafter. 


12. Eingang. In dance music, and less frequently in other 
forms, it is customary, before introducing a new theme in a different 
key, to prepare the way for this change by means of a short modu- 
lation. This is sometimes marked eingang, entrance. It is nearly 
always isolated from the principal strain, and being of an adventi- 
tious character it is played ad libitum. The following excerpt from 
Grieg's Valse Caprice is an excellent illustration. It occurs at the 
end of part n as a means of returning to part I in C-^L minor: 

Ex. 120. 


Op. 37, I. 


The stationary tone above serves to connect the two parts, while 
the harmony descends chromatically until the new mode is suffi- 



ciently prepared. At the beginning of the eingang the rhythm of 
the previous strain is imitated. The correct interpretation is indi- 
cated in the printed copy. 

Rubinstein, in his Tarantella, op. 6, changed the mode from 
B-minor to B -major by introducing the large 3d and large 6th into 
a descending scale figure : 

a tempo. ^^f^ 

pocorit. I .. I I,, I 

Ex. 121. t^JtJ 

The key-tone remains unaltered, but the major mode is anticipated 
by means of the short eingang here quoted. It is therefore some- 
thing apart from the antecedent and consequent periods and 
demands special treatment. With the change of mode there is a 
corresponding change of mood, and this is to be foreshadowed by 
a deliberate performance of the eingang and a perceptible accent 
upon the altered notes already referred to. 

Eingange are frequently introduced into sonatas and other works, 
but their identity is less apparent than in the dance form. We 
must known whether a transition passage forms part of an organic 
whole, whether it is part of the period in which it occurs, or merely 
an anticipation introduced for the express purpose of anticipating 
a particular key. In the first movement of a sonata, for example, 
the second subject is supposed to be in the dominant (or some 
parallel key), and composers usually prefer to connect the two sub- 
jects by means of transition matter. Sometimes this is so closely 
interwoven with the first theme as to offer no suggestion of eingang, 
and in such instances the performer must not endeavor to produce 
an effect which has no underlying cause in the music. But where 
the principal theme is concluded in the original key, and an inde- 
pendent modulation is made to the key of the second theme, this 
modulatory section may be treated as eingang. In the first allegro 
of Beethoven's op. 2, I, the modulation from F-minor or A-flat- 
major is included in the principal theme as an extended period. 
Therefore no eingang appears here, nor in the following : cp. 1 3 ; 

9 6 


op. 14, I and II; op. 27, II. But in the op. 2, II, the nine 
measures before the second subject may be treated as eingang. See 
the modulation marked rallentando. The Eg. includes four meas- 
ures of a tempo. 

A simple example occurs in the finale of op. 2, I, where the 
three forte chords are introduced to prepare the ear for the middle 
part in A-flat, which follows : 

x 1 22 i 1 

f , Eingang. 

~ - b ?- L ^^- ^~ 


ad lib. 

After the period ends in C-minor there is a considerable interval 
of silence, and the meaning of the dominant seventh chord, re- 
peated ff, can not be mistaken. 

An eingang somewhat analogous to this occurs in the largo of 
op. 7. The first theme ends in C ; the second theme is in A-flat ; 
and the two keys are thus connected : 

A complete period ends at (a) ; the eingang begins at (b) ; the 
second theme begins at (c). Every note of the eingang is to be 
played deliberately (especially the last group) and slightly rallen- 
tando. Another instance may be found in the Allegretto of op.. 


10, II. At the end of part n in D-flat there is a silent measure 
and then a simple eingang of six measures leads to part i in F- 
minor. The principal accent is to be placed upon C, not alone 
because that sound must continue, but on account of its dominant 

13. Intermezzo. This term is here applied to intermediate 
passages in a rondo or other form, not to a complete movement. 
According to the smaller definition intermezzo corresponds to the 
interlude in a ballad and serves to relieve the monotony of a 
frequently repeated principal theme by presenting a contrast to it. 
The intermezzo is usually of irregular construction, in which 
uniform phrase divisions are more often absent than present. 

There are several intermezzi in the rondo by Beethoven, op. 51, 
I. The principal one contains five measures, and this requires 
strict movement, with mensural rather than rhythmic accent. . 

The Melody in F by Rubinstein contains two corresponding 
intermezzi. The first descends chromatically one octave ; the 
second ascends one octave in like manner. The chromatic notes 
would seem to indicate a transitional tendency, yet the intermezzo 
ends where it began, on the dominant. No attempt should be 
made to subdivide these intermezzi into phrases. The composer, 
himself, did not do so. 

There is another, and generally more important, species of inter- 
mezzo which serves a twofold purpose : as relief to a recurring 
theme (or to regular periods) and as a means of catenating two 
parts located in different keys. Such an example may be found 
in the finale of Beethoven's op. 10, II. The intermezzo is first 
used to connect the strain in A-flatwith that in D ; afterward there 
is an intermezzo leading from D back to F. In op. 2, I, last move- 
ment, there is an intermezzo leading from the middle theme in A- 
flat back to the prestissimo in F-minor. Also see the rondo in op. i 
14, I. There is an intermezzo of nine measures between the repe- 
tition of the main theme. 

The thematic intermezzo is more free in construction and natu- 
rally corresponds to the impromptu character of interlude. See the 
intermezzi in von Weber's Rondo Brillante, op. 62. 

14. Cadenza. The general character of this embellishment, 

9 8 


and the conditions under which it appears, indicate a more or less 
ad libitum style of performance. Even the measured cadenza (a 
tempo) seldom forms an integral part of the whole, and the main 
idea is therefore arrested in its progress. The simplest examples 
are those which terminate a movement^ Where the ornamental 
passage is written in small notes, and without the dividing bars, in- 
experienced players are usually in doubt as to the proper mode of 
procedure. Rubinstein's La Melodia presents such an instance 
at the end of the coda. There are sixteen small notes included 
within two vertical bars. If these be divided into four measures 
the rhythm will be preserved and a degree of affinity thus main- 
tained : 


On the other hand, the groups naturally fall into triplet figures, 
and if this suggestion prevails it will be necessary to perform a 
triplet to each quarter-note beat. At the close the last four notes 
may be played as even eighths. In connection with the rallentando 
the discrepancy between six and four would not be objectionable, 
thus : 


The accentuation should be light, merely enough to suggest the 
mensural outlines indicated by the added bars. Notwithstanding 
the ascending form of the cadenza, it should produce a vanishing 



effect, like an isolated cloud which is dissipated by the air and so 

Cadenzas vary in length from one measure (as in vocal and violin 
music) to the extensive and elaborated addendas written for the 
classic concerti of Mozart and Beethoven. In the first movement 
of Mr. Ad. M. Foerster's trio, op. 29, there is a cadenza for the 
piano of 36 measures. It consists of a paraphrase of the main 
subject and was marked by the composer quasi ad libitum. 

A cadenza to Chopin's F-minor concerto has recently been 
issued, composed by Rich. Burmeister. 

There are several ad lib. cadenzas in Mozart's Fantasia in D 
which it would be well to analyze. The first occurs after a section 
of the main theme in A-minor. There are four irregular groups of 
mixed scales descending, and each of these should be executed 
very nearly in the time of a quarter-note beat, moderate tempo. 
The four diminished chord groups form themselves naturally into a 
full measure. The final E-flat is to be accented (as though it fell 
upon the beginning of a measure) and sustained by means of the 
damper pedal. In the second cadenza mensural proportion disap- 
pears, though each of the groups beginning with B-flat should be 
distinctly marked. At the close there is a chromatic scale passage 
ascending and care must be exercised in connecting this with the 
principal theme in D -minor. A, being the dominant, should be 
rather prominent throughout. The following reading is recom- 
mended : 

poco rit. ^*-^^"" 

Ex. 125. 

^^- J 

This leads naturally to the melodic note,/, and serves to restore 
e mensural equilibrium which was disturbed by the cadenza. 
In part n there is another cadenza ; and in the last of this there 
an attempt at representing, by means of notation, a rallentando 
ect. The four concluding groups of thirty-second notes, one 


group of sixteenth notes, and one group of eighths. If interpreted 
literally this would scarcely express the composer's idea, which was 
a gradual lengthening of the quick notes with which the cadenza 
began, thus : 

Ex 126. 


poco a poco rit 

a tempo. 

To play not only each group, but each note, a shade slower than 
the preceding is not a very simple matter, but it is what Mozart 

Where the ornamental groups are arranged in regular measures, 
as in Paderewski's Krakowiak, op. 9, III, the design appears more 
plainly than it does in such instances as were quoted from the 
Mozart fantaisie. 

To these points it is necessary to add only this, that the cadenza 
is more or less episodial and does not require the same treatment, 
either rhythmically or melodically, as do the regular strains. A 
certain ad libitum style should be observed. Sometimes the move- 
ment is modified or arrested ; at other times the mensural accents 
are omitted. Alternate crescendo and diminuendo, and brief pauses 
upon certain tones, also play important parts. (See the close of 
Chopin's Nocturne, op. 9, II.) 

15. Passage. This general term is applied specifically to a series 
of chord or scale figures in thematic style, and of irregular period 
construction. The simplest examples consist of a sequence con- 
tinued beyond its natural length, thus diverting the attention from 
a regular rhythmic arrangement by phrases. Passage occurs most 
frequently in "free fantasia" or development; also in preludes, 
intermezzi and terminations. An example is quoted from the " per- 
petual movement " rondo. It occurs in the termination and con- 
tains twenty-seven measures. The passage at first consists of a 
series of diminished seventh chords descending and ascending. 
This is followed by a continued sequence on the four-six chord of 
C, leading to the last recurrence of the rondo theme. The entire 



passage is irregular in construction, and neither the chord figures 
nor the sequences can consistently be divided into phrases. Differ- 
ent conditions are here presented, and therefore a different mode of 
treatment is required. Nearly all passages are transitional and rep- 
resent more or less of emotional excitation. The one under notice 
is a fair illustration, and part of this is quoted : 

Ex. 127. 

Von Weber. Op. 24. 


The melodic outline here continues to ascend until it reaches A, 
which is the climax. Then there is an anti-climax at the end of these 
diminished seventh chords. Considerable reserve force is there- 
fore required, and this is the principal difficulty. All of this ter- 
mination is marked^", but that is evidently a very general direction, 
since it would, if literally applied, destroy the climacteric effect. 
The performer would better commence the passage m /and then 
increase it to^\ 

Another interesting example of passage may be found in the first 
allegro of Beethoven's op. 53, known as the " Waldstein Sonata." 
Reference is made to the pedal passage commencing with measure 
142 and extending to 1 56. Beginning pianissimo, there is a gradual 
crescendo to fortissimo, immediately before the return of the prin- 
ipal theme in C-major. In von Bulow's edition the style of per- 
formance is carefully indicated. 

From Mozart's piano sonata in C (XV Peters' Ed.) an inter- 
sting quotation is made : 


Ex. 128. 



(II. Litolff'sEd.) 

^r* 1 

This occurs in the development and leads to the reprise (b). The 
sequence figures are to be performed connectedly, with a vibratory 
accent upon the bass notes c, b, a, g-sharp and a secondary accent 
upon the corresponding melodic outline above : f y e, d, c. This 
will reveal the motive above as well as that below. 

The harmony makes the cadence so naturally that only a slight 
rail, is necessary. Observe the dominant relation in the passage : 
A D, G C, F B y E A ; also the melodic outline above, 
already referred to. 

16. Appendix. This is a species of codetta and belongs pri- 
marily to vocal and concerted music. It is used to fill in a void 
between interrupted sections of melody, or to connect one strain 
with another by means of a brief modulation. " Adelaide," or 
" Ah ! Perfido," by Beethoven ; " Spring Night," by Schumann ; 
" La ci darem la mano," from Don Juan ; in fact almost every fine 



song affords an illustration. See end of first period in Kucken's 

" Good-night, Farewell." An example is quoted from Dudley 
Buck's " Ave Maria " : 

Ex. 129. 

_ _ _ (a). (b). ( 





i i- -f^i ?' 






In the vocal part a strain ends at (a) and the first theme is resumed 
at (b). The appendix in the accompaniment serves two purposes : 
It fills the void occasioned by the rest in the voice part, and 
connects the two periods by modulating from D-flat back to G-flat. 
In applying this device to instrumental music it will be necessary 
to consider the appendix as issuing from some other instrument 
than that which delivers the main theme. This separate view of the 
matter may be facilitated by performing examples like the last as a 
piano solo, but with due regard for the fact that it consists of a 
lyric melody and an accompaniment. 

An instrumental illustration is taken from Lassen's ballet music, 
" Love Above all Magic " : 
Ex. 130. 


*-M = F - * 

?^ ESSE 


The first period, in G, as well as the second, in B minor, contain 
these appendixes. They are here performed like echoes, though 
the appendix may appear below as well as above the principal 

Innumerable instances similar to the following might be quoted : 

Ex. 131. 

rit. > 

kJL I 


The appendix of one measure and a half occurs between the 
ending of the first period at (a) and its repetition at (b). See the 
lower slur. 

17. Refrain. The refrain sometimes corresponds to episode 
and at other times resembles the burden to a song. " Flowers," 
by Bradsky, may be taken as an illustration. At the end of every 
verse there is a cantabile strain which chants praises to the rose. 
This is the refrain. " The Old Song," by Lassen, is somewhat 
similar ; so is the " Three Fishers," by Parsons. 

Jean Nicode, in his "Ball Scenes," op. 26, introduces at the 
close a refrain of singular construction. The waltz proper suggests 
a series of ball room pictures, but in the langsam (a grand prome- 
nade) and the Refrain which follows, all is changed. Observe how 
the latter reflects the mood of one dependent upon a smile. The 
tranquil nature of the dissonances, the unsettled tonality, and espe- 
cially the restraining effect of the sustained D against the E-flat 
above, all point to an illusive denouement and a repressed passion. 
The Refrain is indicated by the composer. (See the original four- 
hand arrangement.) 

The next example is quoted from Ph. Scharwenka's " Dances 



Polonaises," op. 38, II. The refrain occurs after the end of part 
I in B-flat and before the second theme in G-flat : 

Ex. 132. 

REF. Lento. 




This bit of shadow serves partly as an intermezzo and partly as an 
epitome of the work, which tells of smiles and tears. All the 
instances quoted are either ad lib. or un poco piu lento. 

18. Episode. This term has been used indiscriminately in 
musical criticism. But to the author it seems desirable that the 
word episode should be confined to its specific designation. The 
country dance introduced by Beethoven into his pastoral sym- 
phony may properly be termed an episode. The sudden change 
from three-fourth to two-fourth measure, and the unmistakable 
dance rhythm with its grotesque accents, are not such features as 
we might expect to encounter in a symphonic work. Yet, as there 
was a merry-making scene among the peasant-folk it is reasonable 
to suppose that they indulged in a rustic dance. In the present 
instance the style is sufficiently indicated, this being impression 
music. But as a general principle the author confines himself to 
the remark that episode indicates merely a digression, or the intro- 
ducing of adventitious matter. It is therefore a notice to the per- 
former that some separate incident has been introduced, and that 
its import must be apprehended. 

Examples somewhat similar to that cited from Beethoven occur 
in Raff's " Im Walde," H. Hofmann's " Frithjof," and in Gold- 
mark's symphonic suite, " The Country Wedding." Piano and 
organ compositions contain fewer examples of episode. In the 



" Elegy to Night," by Jambor, there is a short episode as middle 
part, piu mosso. 

Another instance may be found -in the Scherzo by Chopin, op. 
31. The episode begins thus : 

Ex. 133. 

Sotto voce. 

This tells its own story and needs no words of comment. 

19. Continued Thesis. This is fully illustrated in Complete 
Musical Analysis and it is only necessary here to deduce from it a 
mode of procedure for practical performance. 

The continuation of thesis usually rests upon a harmonic basis 
which is so conceived as to form a continuous chain without resolv- 
ing cadence. A melodic sequence also may be repeated in such 
manner as to constitute what is here termed continued thesis. In 
either case there are two important features to be considered : I. 
The unity of design ; 2. the increasing tension and exaltation 
which the continuation of thesis usually expresses. 

(There is, apparently, considerable similarity between passage 
and continued thesis, but a distinction ought to be made between 
them. Passage work is more thematic and usually more artificial. 
Continued thesis is essentially melodic and forms an integral part 
of the music structure. See Ex. 1 34. Passage resembles extended 
sequence work, or certain kinds of cadenza ; the continuation of 
thesis forms part of an organic whole and represents a central 

The first quotation is from a " Night Elegy " and occurs in the 
termination, thus : 


Etig. Jambor. Op. 23, IV. 

Ex. 134. 

Andante con moto. Appassionato. 

ft- * * I f^ 1 " ^ I _ -0- ^"^ |^^ 

^J 1 . I Mm. ^ LLJ r*. 3 


--- E= 

The sequences in the uninterrupted melody are first to be ob- 
served. Notice the peculiar manner in which the period is extended, 
and how the phrase divisions are lost in the passionate repetition 
of the first sequence figure. Then observe the chain of harmonies 
in the accompaniment. Though the middle parts proceed some- 
what in a dominant relation (as .Fto B-flat, G to C), the continu- 
ous theme above and the pedal -note below serve to maintain the 
interest and to counteract the effect of an intermediate cadence. 
The expression becomes more and more impassioned until the 
climax (ff) is reached. No cadence occurs here, however ; but 
after the diminished chord on E natural is resolved to F-minor the 
emotional excitement begins to subside, and finally ends with a 
perfect cadence on A-flat. The continued thesis consists therefore 



of eleven measures, and ends with our quotation. The manner of 
performance is indicated (as well as such moods may be) by the 
words and symbols in the example. 

The next quotation is more extended but less ebullient than the 
previous one : 

E. Jambor. Op. 23, IX. 

From this point the thesis continues during twenty-one measures. 
This cleverly prolonged period will justify itself without any pre- 
arranged plan of phrasing, so-called. The main difficulty will con- 
sist in performing the entire twenty-one measures as an integral 
whole, a Continuous and connected period. A carefully graded 
and well-sustained crescendo is particularly desirable here. The 
thesis begins piano and the termination is marked fortissimo. To 
gradually increase the volume of tone to this latter degree is the 
principal task for the performer. 

Rubinstein's Tarantella, op. 6, contains an interesting illustration 
of continued thesis. It commences before the figure of the intro- 
duction recurs and includes the twenty-one measures of intermezzo 

Beginning thus, 



Ex. 136. I^Kiff 

it continues during thirty-seven measures. The impression of 



as dominant pedal-note throughout this long, unresolved 
passage considerably simplifies the performer's task, because there 
can be no cadence while the effect of dominant harmony prevails. 

The principal requirement in performing a prolonged thesis is, to 
maintain the continuity of design, whether it embodies a highly 
wrought emotional expression (as in Isolde's " Love-death "), 
vaguely defined longing (Jambor, op. 23, IV), or the realization of 
hope deferred (Mozart fantasia in D, and Jambor, op. 23, IX). 


20. Carillon. Bell motives and carillons have become free 
possessions to modern composers, though only a limited number, 
apparently, have been used. These bell motives frequently have a 
special significance when employed in certain works, and therefore 
demand consideration from the performer. In North America we 
know but little of carillons, change-ringing, curfew bells, matins, 
etc. But in the older countries bell casting is a fine art, and the 
carilloneur devotes a lifetime to his profession. The bells range in 
weight from ten tons to a couple of pounds, with corresponding 
varieties of pitch. In religious rites and services the use of bells is 
particularly significant. 

For secular purposes certain popular tunes have been transferred 
to the carilloneur, and the combinations in change-ringing have 
developed several motives which are characteristic. They may be 
called bell music. 

In the opera, " Chimes of Normandy," one of these motives is 
effectively employed as an accompaniment to Germain's solo, " The 
Legend of the Bells" : 

Ex. 137. 

Also small bells are employed for musical purposes. Something 
of this nature is suggested by the well-known bravura piece, " La 
Campanella," Paganini-Liszt. 

The most satisfactory carillons are those that are evolved from 
the natural cadence harmonies. An analysis of these shows their 
common origin. The first example (a) may be considered a model : 




Ex. 138. 

Every tone of the major scale (excepting 7) is here employed ; yet 

the motive naturally rests 

upon : Ex. 139. 

For obvious reasons the chord motives are most satisfactory. A 
few of these are presented here : 

Ex. 140. 

/1\ / \ / 1 \ 

. (b). j . (c)^ ^ , i (d). i 

One of the most curious instances of the artistic application of 
bell motives is the rondo, " Midi," by John Field. It is not one 
of those inane effusions in which an occasional monotone is the 
only justification of the title, but a charming genre piece conceived 
from first to last in the carillon style and developed entirely from 
bell motives. The principal rondo theme is this : 

Ex. 141. 

This is a variant of the motive (d) Ex. 140, afterward used more 
prominently in its original form. Even in the intermezzo (16 to 25) 
the tintinabulations continue. The second subject (in )is founded 
upon a bell motive exactly like our Ex. 138 (a). This ground-base 
serves as counter-subject to the theme above, which likewise is in- 
cluded among the carillons. (Gounod employed a scale figure like 
this in his song, The Angelus.) In the present instance the C. 
S. is quite as important as is the upper theme, and wherever these 


bell notes appear they should be treated somewhat in the manner 
of leit motif. During the continuance of this part the upper theme 
is slightly varied and resembles our motive (d). 

After a brief intermezzo the carillons in B are repeated. Then 
the main theme recurs, and this is followed by an episode in 
E-ininor which seems to ff 

suggest the clanging of 
fire-alarm bells, thus : Ex> 142 ' 

These were marked by the composer marcato, and they recur in a 
modified form in the sequences beginning on A-flat. 

The coda very appropriately represents the natural termination 
of these carillon experiences : The bell of a tower clock begins to 
mark in slow monotones the hour of twelve, while from distant 
belfries we hear the accompanying chimes. The different parts 
stand in the following relationship, dynamically : i. The monotone 
of the hour bell is to be most distinct. 2. The carillon previously 
used as C. S. 3. The chime of small bells in the highest part, 
which requires the least accent. 

And now must be given a few directions with regard to the 
manner of producing bell effects. Without entering into an ex- 
planation of the material composition of bells (nearly all metals are 
used), it may be stated as a fact that they usually emit an imperfect, 
composite tone. The vibrations are necessarily of a metallic quality, 
and the impact of the mallet (or clapper) against the sound bow 
imparts to bell tones the character of all percussive instruments. 
The piano lends itself naturally to this effect. The vibratory quality 
is produced by sounding the bell note with a staccato touch con 
pedale. And as a matter of fact all carillon effects require a very 
free use of the damper pedal in order to imitate the more or less 
unmusical vibrations which result from the bell sounding in dispro- 
portionate parts, instead of vibrating as a whole. A few cam- 
panella effects are indicated in the following : 


Ex. 143. 

W. G. Smith. Op. 56. 



The principal theme is played legato and the chord well sus- 
tained, while the 1. h. passes over and lightly touches the bell notes. 
If these are played demi-staccato, and both pedals are pressed 
immediately after the arpeggio chord is sounded, the effect will not 
fail of its suggestiveness. The soft pedal is here indicated (U. C.) 
not so much for the purpose of subduing the tones as for changing 
the quality ; because the bell notes are a prominent feature of the 
composer's sketch, and yet they must be plainly distinguished from 
the regular theme. In the repetition of this initial period a more 
melodious bell motive is introduced as C. S. "This requires differ- 
ent treatment : 

Ex. 144. 

94 94 

-<- ^ ' 




The author has indicated a slurred staccato style for the B. M. 
in order to properly distinguish it from the theme below. 

Attention is now directed to the magic fire scene from " Die 
Walkure," arranged for piano by L. Brassin. In the original score 



Wagner introduced the glockenspiel to reinforce certain melodic 
notes. He did not intend to produce a simple bell effect, but to 
impart to the musical delineation a certain piquancy and sharpness 
of outline. In the present instance it is desirable to create a very 
bright effect by means of energetic staccati, especially in these rein- 
forced tones : 

Ex. 145. 

*-"-"- \-j~-* 

5=p^,_5 f*f:-f 


After having listened attentively to the" original, it will be com- 
paratively easy to make a fair imitation of the intended effect by 
means of Brassin's clever transcription. 

21. Ground-Base. This device consists of the repetition of cer- 
tain cadence harmonies, or any figure which may serve as foundation 
to a varying theme above. The ground-base is frequently a constitu- 
ent element of the chaconne, the pifferare, and other forms. Handel's 
chaconne in F, and the more famous one in D-minvrby Bach,* afford 
good illustrations of the artistic application of ground-base. A more 
modern example is quoted from F. Hiller's " La Ronde de Nuit" : 

Ex. 146. 

Con moto. 

Op. 146. 

The themes are contrived with so much ingenuity that they do 
not seriously interfere with the ground-base, which continues 
throughout the rondo. The constant repetition of the base figure 

* This was composed for violin alone, but it has been arranged by Raff as a piano duet. 


shows that it is a prominent feature of the music, and the occasional 
dissonances which result are perfectly musical, because the design 
renders them necessary. 

The second valse by B. Godard contains a ground-base figure 
which continues throughout the waltz. These cadence harmonies 
(especially the real-base part) are to be treated as counter-theme 
to the figurations above. 

Usually it is the general harmonic effect of the G. B. which is 
intended to color the musical sketch. This is true of the examples 
thus far quoted. Other instances are mentioned in the Compendium. 

22. Pedal-note. The various uses to which pedal-note has 
been applied by modern composers show the importance of this 
permanent tonal foundation. It may serve as connecting link in a 
chain of harmonies ; as foundation for a continued thesis ; as 
equilibrium in cases where the inverted chords are too much out 
of balance ; it may add tenacity to a design requiring this quality, 
or it may contribute seriousness to an otherwise trivial passage. 
One of the most remarkable instances is the Ballet of Sylphs 
from Berlioz' " Damnation de Faust." The entire movement rests 
upon a tonic pedal sustained continuously by the violoncello. 
While the fairies trip their mystic measures the softly sustained 
organ-point helps to picture the peaceful and quiet scene, while the 
unhappy Faust sleeps and dreams of bliss. 

Liszt wrote a piano arrangement of this Sylph ballet, but since 
then a sostenuto pedal has been applied to the piano by Dr. 
Hanchett, of New York, and this attachment can be made very 
useful in prolonging an organ-point without interfering with the 
-other pedals. In the Sylph ballet there are three parts to be treated 
independently, and their relative degrees of importance are in this 
order : (i) The principal theme ; (2) the abbreviated chord figures 
in the middle ; (3) the pedal-note. The manner of performing this 
latter will depend upon the quality of the instrument used. If the 
base is sonorous, the P. N. may be made to vibrate during eight 
leasures without being retouched. But in no case must the key 
>f the pedal -note be pressed oftener than once in a phrase of four 
Leasures. The finger which touches the organ-point should be in 
:ontact with the key before the time of percussion arrives, for eyen 



the slightest evidence of mechanical action is to be scrupulously 
avoided throughout this movement. 

The P. N. as foundation of a continued thesis has already been 
illustrated. By referring to that example (134) it will be observed 
that the organ-point is a prominent feature of the passage, and 
must be well sustained and sonorous. 

The next quotation is from the final coda to a serenade for string 
orchestra (arranged as a piano duet by the composer) : 

E. Elgar. Op. 20. 

( Continued in repetition an octave lower. ) 

The nature of the harmonies is such that if they were accom- 
panied by fundamental bases the effect would be trivial and incon- 
gruous ; whereas the tonic pedal imparts to the music an air of 
repose that is charmingly satisfactory. The organ-point is to be 
scarcely more than perceptible. 

The following double pedal illustrates an opposite phase of this 
subject : 

Ex. 148. (a). 

Ph. Scharwenka. Op. 38, II. 

t=p=p tfd=p^=l=Q 





H K I- 




The principal theme appears at (a) in full harmony with a some- 
what rigid, ponderous accompaniment. The double pedal in this 
instance expresses boldness and tenacity of purpose, and is there- 
fore to be performed in an unyielding manner. The consequent 
period at (b) is much more darkly colored by means of the chro- 
matic harmonies. These are somewhat out of balance when consid- 
ered fundamentally. The dominant pedal-note, however, serves to 
maintain the tonal poise and is absolutely essential to the design. 

23. Drone Base. This is mostly an imitative effect, as illus- 
trated in the musette, tambourin, and other country dances. In 
the Scotch symphony Mendelssohn introduced a drone base as 
suggestive of the bagpipes, thus imparting a local color to the 

Where the drone base in the musette is written continuously by 
means of ties it is to be re-touched at certain regular intervals of 
time in order to keep the pedal -notes in vibration. (The musette 
was incapable of forte effects, and therefore the dance of that name 
is almost invariably quiet.) A brief quotation is presented : 

E. d' Albert. Op. 1. 

Ex. 149. 



No special accentuation of the drone base is here necessary, 
because that would give to the music a syncopated, incongruous 
effect. But the musette imitation may be suggested by prolonging 
the lower d in this manner : 

In addition to the harp pedal the damper pedal also is recom- 
mended in order to continue the vibrations, for it is to be observed 
of all these bagpipe instruments that their tone is directly opposite 
to that of pulsatalic instruments like the piano. Something of an 
organ tone is consequently desirable in music of this character. 

The double drone (tonic and dominant) requires different treat- 
ment, especially where the notes are tied, as here : 

Ex. 150. 

Wilson G. Smith. Op. 34, II. 

Alia musette. 

jLlln musette. . , NMMH^H 

1:2=^ ^=r++^q-i:T::tg= ^^4= 


m &-*- m~\-& 

r5 rfr 

The base is accented here in order to make it vibrate during two 
measures. Otherwise the drone is soft and should attract very 
little attention. In the repetition of this period a peculiarity occurs 
which may need explanation : 

Ex. 151. 

The double drone should continue to sound as at first, for it is 


not to be presumed that the bases stop singing merely because the 
tenor part has a bell note on the dominant. By means of the 
sostenuto pedal this effect may be produced : 

Ex. 152 

8. P. 

The composer would doubtless have notated the passage in this 
manner had he written it for band or orchestra. 

Another interesting example may be found in the musette to F. 
Dreyschock's E-major gavotte. (Part I is properly a bourree.) 

24. Recitative. The general principles governing vocal recita- 
tivo apply in a limited sense to instrumental music in this style. 
Having no specific text to follow, the instrumentalist is driven to 
the necessity of imagining such words as seem to apply to the 
nature of the sounds. With this supposed text before him the 
performer will experience less difficulty in delivering the recitativo 
in an impressive manner. All that can be stated a priori is this : 
That recitando passages should be delivered in a broad, declama- 
tory style, without strict regard to the rhythmic value of the notes 
or regularity of the movement. The " Danse Macabre" contains 
an interesting example. In the midst of this ghostly revelry the 
horn sounds a note of warning, all the dancing and fiddling sud- 
denly stops, and the chanticleer is heard signaling the approach of 
day. In the words of the composer, " Death utters a mournful 
declamation," and then the participants in this noisy incantation are 
supposed to disappear with the darkness which called them forth. 
The declamatory passage is as follows : 

Ex. 153. Saint-Saens. 

, Recit. 

f - ^_ 



These notes give but an imperfect indication of the intended 
effect, for surely we cannot measure them out in this manner. 
Perhaps the following arrangement may be of assistance : 

Ex. 154. 

ad lib. 

Of course, it is to be understood that the regular movement 
(allegro) is to be relaxed during this declamatory passage. The 
accompaniment consists of softly sustained chords below, and these 
offer no obstacle to the ad lib. style of the recitativo. 

The introduction to Liszt's second Hungarian Rhapsody is a 
recitativo. To attempt to produce the proper effect in this intro- 
ductory declamation by maintaining the strict movement (as some 
do) is to attempt the impossible. " Counting the time " should be 
entirely dispensed with during the recitativo, and the non-legato 
style is generally appropriate. In truth, a slight disconnectipn of 
the tones, with distinct articulation, serves to produce the effect of 
actual declamation, as in parlando passages the experienced singer 
will abandon the vocal style and seek to emulate the orator. If, 
for example, the student will execute the recitativo from " Saint- 
Saens " (Ex. 153) with a single finger, he will approximate the 
manner in which certain declamatory passages ought to be per- 

25. Coda. This term has been variously applied, and as a con- 
sequence it is not clearly understood by young performers. In 
fugal parlance the word " coda " refers to a few supplementary 
notes (sometimes less than a measure) which may be added to a 
subject in order to admit the response according to tonal require- 
ments. Such instances should be called by the diminutive term, 
codetta, or appendix. The simplest examples of coda (tail) are those 
which correspond to postlude in a song, and which are introduced 
after the movement has apparently ended. In such instances the 
coda is separated from the preceding and appears as an after- 
thought. When built upon the principal motive the coda is gener- 



ally ad lib. in style. Such is the character of the coda in 
Schubert's Hungarian Divertisement, op. 54. See the last 
thirty-six measures. This beautiful coda diminishes to the close 
and the movement is considerably relaxed. Similar in style is the 
coda to the last variation in Beethoven's op. 26. See also close of 
the adagio in the same master's op. 27, II. This coda is founded 
upon the principal motive, and the very monotony of grief is 
expressed by the reiterated monotone, g-sharp. Another instance 
occurs at the close of Schumann's charming berceuse, op. 124. 
The last seven measures form a coda, and the style is ad lib. 

The examples quoted (of which hundreds similar in design might 
be mentioned) are short conclusions, somewhat regretful or plaintive 
in character. In conclusion mention is made of the coda in the 
largo to Beethoven's op. 7. It forms an united period with the 
termination of the principal theme. A contrasting example is 
presented in the last of " Au Matin," by Godard. 

It is remarkable, in view of the instances cited, that no coda 
appears in any of the eight lyric pieces, op. 38, by Edvard Grieg. 
The concluding measures of "Asa's Death," from his first Peer 
Gynt Suite, is one of the few exceptions. 

26. Termination. The author applies this term to the finale of a 
movement when the concluding portion is joined uninterruptedly to 
the preceding, as in united period. All other writers call this coda, 
but in actual practice it seems desirable to make a distinction 
between Coda and Termination. As a simple illustration the con- 
cluding portion of Haydn's " Gipsy Rondo" is cited. After the 
final repetition of the principal theme this termination occurs, some- 
what in form of an united period : 

Ex. 155. 

P. U. 

From G-major Trio. 



Observe that the termination is not separated from the previous 
strain (a) but joined to it without interruption. Observe, also, 
that the new matter (Ter.) is of a sprightly, animating character, 
and admits of no ritard, but rather demands an increase in move- 
ment to the close. The author's object in drawing a distinction 
between coda and termination is, therefore, to mark an actual differ- 
ence between one and the other. Coda (especially when brief) 
is usually to be played apiacere and in a slower movement. Whereas 
termination is more or less turbulent or impatient, and in nearly all 
instances is to be performed con anima, or accelerando. 

A more elaborate termination occurs at the end of Beethoven's 
op. 2, I, from this point : 

This begins decidedly, and with increasing animation to the end. 
In the Stuttgart edition this Ter. is marked " close II." 

The andante to Mozart's C-major sonata is another illustration 
of the point under notice. At the end of the romanza a termina- 
tion of eleven measures begins thus : 

Ex. 157. 

The author has marked this cresc., because its impatient charac- 
ter is plainly manifest. (See Ex. 51.) 

Mendelssohn's " Spring Song," op. 62, VI, presents a somewhat 
similar instance. See the last twelve measures. The terminating 
portion of Beethoven's movements is usually more extended. Thus, 



at the end of the rondo in op. 13, after the principal theme has 
appeared for the last time, there is a termination of more than a 
page. The annotate rs have labeled this " Coda," but it would be 
better understood as termination. It is not isolated from the pre- 
ceding nor does it appear as an afterthought. It is impatient, even 
exciting, in its -expression, and calls for increased tone and move- 
ment. The termination referred to begins thus : 

Ex.158. ^ f ^ f 

(Ter.) .^feVf: -^V Et::E * ^-fai. 

1 ', ! ! "T~ ' ' ' ' *~ "*""" ' 5r " * 

i n 

Many examples of termination may be found in works composed 
during the present century. See the last thirteen measures (marked 
accelerando) in Chopin's A-flat Waltz, op. 42 ; also Valse 
Caprice, Karganoff, op. 16. 

27. Recollection. At the close of a movement composers 
frequently introduce a fragment of the first or second subject as a 
reminiscent expression. Thus, in an Adagio beginning like this, 

Ex. 159. 


the composer includes in the last five measures a brief recollection 
as codetta to the movement : 

Ex. 160. 



_= DO: 




The effect is inclined to be regretful and pensive. Similarly 
Mozart employs a brief recollection at the close of his rondo in D : 

Ex. 161. 

This is a variant of the original motive and occurs in the coda. 
The nature of the sounds will, of course, influence the style of per- 
formance, and where the motive is so altered as to sound regretful 
the tempo should be taken more slowly. This is more plainly 
illustrated in the next quotation. The original motive at (a) 
appears in the recollection as at (b) and (c) : 

Kuhlau. Op. 66, II. (4 Hands.) 


The ritenuto is plainly indicated by the dark tones (a-flat and 
e-flat in the key of C-major), and these are to be closely connected 
with the tones which precede and follow them. 

A similar instance occurs at the end of the Andante in Schu- 
bert's A-major sonata, op. 120 : 

Ex. 163. 

See also the last seven measures of Chopin's Nocturne, op. 
37, II. The recollection is in the nature of a benediction. 



28. Stretto. Stretto is here applied to the few last measures of 
a movement when they are impatient or precipitant in style. The 
bolero in Moszkowski's op. 12 contains an example. After the last 
perfect cadence there is a stretto of eight measures, and this is 
played faster and faster to the end. A more characteristic illustra- 
tion is here quoted from Grieg. It is preceded by a brief recollec- 
tion (piu lento) and a pause. Then the stretto takes place : 

Ex. 164. 





The manner of performance is sufficiently indicated by the 

Other examples of stretto will be mentioned in the Compendium. 
But it is unnecessary to add more illustrations here, because the 
stretto is invariably intended to be performed in the manner de- 



As a preliminary step in the study of style and expression, it will 
be necessary to pass in review certain species of the dance form, 
especially those whose characteristic features are of musical value. 
One of the oldest and most important of these is the 

Zarabanda. It is of Moorish origin, but was well known 
to A: Scarlatti, Corelli, Purcell, and Bach. The main characteris- 
tics to be observed in performance are the slow and majestic 
movement and the rhythmic accent, which falls upon the second 
beat of certain measures. The mode is usually minor ; the measure 

* or |f ; the form, that of a ballad dance, two periods. 

Corelli's sarabandes are largo ; those of Bach are andante and 
andantino. The movement varies from ( 1= 76) to ( |= 88). 

One period from Corelli's D-minor sarabande is quoted as a spe- 
cimen example : 

Ex. 165. 

It is seldom that the special accent is required in every meas- 
ure, but where the note which falls upon the second beat is a half 
or a dotted quarter, the rhythmic accent is intended, though it may 
not be otherwise expressed. In the sarabande from Corelli the 




.accompaniment is mostly in this rhythm, 
the special accent. 

In Bach's D-minor sarabande the rhythm is 

the one in A-minor is similar * * * f 

, which implies 

* * * * 

minor sarabande begins, 


Corelli's - 

The accent symbol 

does not always appear, nor is the movement always indicated out- 
wardly ; but the performer is supposed to know the characteristic 
features of these old dances and to be governed thereby. 

Corrente (Coranto). This is an old Italian country dance of 

.a lively character in 5. measure. It begins upon the last half of 

the third beat, which gives to this particular part of certain meas- 
ures a rhythmic prominence. Not all correntes have this peculiar- 
ity at the outset, but the fact remains that composers have recog- 
nized it, and in the course of the music a note will appear upon the 
final half beat as the beginning of a phrase or a rhythm. This 
point is illustrated in the F-major corrente by Handel : 

Ex. 166. 

Each semi-phrase is isolated from the last eighth note in order 
to mark the rhythmic divisions of the corrente. The slurs were 
placed by Von Biilow, who, in a marginal note, states that this 
dance "generally commences upon the last eighth of a measure." 
If the last note at the end of the slurs is played staccato (as writ- 
ten) it will serve to isolate the termination of the scale passages, 
and thus obviate the necessity for a special accent upon the sixth 
eighth note. The cadences also end upon the fifth eighth note, 
which, as Von Biilow observed, " serves to illustrate the manner in 



which the repeated parts [periods] are connected with each other,' 
thus : 

Ex. 167. 

The first period (a) here terminates upon <r, and the last eighth 
note belongs to the second period (b), not to the measure in which 
the e occurs. 

Numerous examples occur in the harpsichord suites and partitas 
and most of these begin upon the last eighth note, or the last six- 
teenth in 1J measure, as in Bach's E-minor corrente from his sixth 


Gavotte. This popular movement was not thoroughly devel- 
oped until the latter part of the seventeenth century. Hence 
we find examples from Dupont, Lock, Lully, Corelli, and even 
Rameau, which only approximate the gavotte form and rhythm of 

a later day. The gavotte is really a moderate movement in 3t 

measure, with regular rhythmic divisions, beginning upon 3 and 

ending upon 2, thus : 

Ex. 168. 

J. S. Bach. 

Since rhythm is the musical expression of action and motion, it 
becomes the mainspring of all dance music. But the gavotte does 
not require strongly marked accents, and the punctuations between 
all well-defined divisions obviate the necessity for special accentua- 
tion. In the irregular or extended periods, such as occur in this 
gavotte, from the sixth English Suite (second period of part I), the 
normal accents may be observed as a relief to the regular gavotte 
rhythm. When the dance is divided into two parts, a brief pause 


I2 9 

should be observed before a new part begins. This is all the more 
necessary where the key or mode is altered in part II. 

Musette. This is the same in measure, movement, and general 
rhythmic divisions as the gavotte. In keeping with its name, the 
musette is founded upon a pedal note or upon a drone base of tonic 
and dominant combined. Another peculiarity of the musette is 
that it is lighter in character and more quiet than is the gavotte. 
When the former is added to the latter, as part II, it is proper to 
pause briefly at the end of the musette and then repeat the gavotte. 
In the D. C. play each period only once. 

Bourre'e. This is a French dance, older and livelier than the 

gavotte, which it somewhat resembles. The measure is ^f-, but the 

rhythmic divisions begin upon 4 and terminate upon 3. Frequent 

punctuations are therefore required between the third and fourth 
beats of certain measures : 

Ex. 169. 

Bach. 2nd English Suite. 


A more pointed staccato may be included at the places marked 
X, but the rhythmic divisions occur here so regularly that very 
little punctuation is necessary. 

Rigaudon. In some of its features this is similar to the bourree. 
Both are in common measure, and both commence upon the last 
beat. But the rigaudon is more rustic in style and the movement 
should be alia breve. It is, in fact, a rollicking country dance and 
ought to be played with considerable rough vigor, suggesting 
bumpkins and hobnails. There is an excellent rigaudon in Grieg's 


" Holberg Suite," op. 40, two or four hands. (The original is for 
string orchestra.) 

Allemande. It is not here necessary to inquire into the origin 
of the allemande (Fr. allmaui), especially since that is a matter of 
doubt and dispute. It is an old German dance, and originally con- 
tained the leaping step. Nearly all composers of the harpsichord 
epoch employed, the allemande in their suites and partitas. Each 
of the six French suites by J. S. Bach opens with an allemande. 
They are in common measure, allegro or allegretto, and almost in- 
variably begin upon the last sixteenth note. This gives an impor- 
tant clue to the phrasing. The leaps (extended intervals) are also 
prominent features, and must be duly recognized in the perform- 
ance. An example is quoted from the sixth French suite : 


Ex. 170. 

_r_J etc. 

One of the most interesting allemandes is that from E. d' Albert's 
piano suite, op. I. (The counterpoint is worthy of Bach.) A 
fragment of this is quoted to show the connection between the pre- 
liminary note (a) and the termination of certain rhythmic groups : 

i I I ^H h-j \-r*-7W | 

The entire passage is slurred, and therefore the reading depends 
mostly upon rhythmic accent, as indicated in the example. It is 
worthy of note that this allemande contains the same characteristic 
features which were recognized by Couperin and Bach, thus prov- 


ing that the allemande is something more than a thematic exercise 
in sixteenth notes. 

Through carelessness or ignorance pianists too frequently ignore 
the individual peculiarities of these old dance movements and sac- 
rifice thereby many legitimate artistic effects. What is still more 
serious, they fail to reproduce the spirit of an important bygone 
age, which lies dormant between the pages of those antique suites 
and partitas. These embrace an epoch beginning with Fresco- 
baldi and Lulli, and terminating about 1750. 



Minuet. This is classed in the Mozart epoch because the 
earlier examples were not clearly defined in respect of those quali- 
ties which characterize the finest specimens of minuet. Composers 
of the seventeenth century occasionally introduced into their suites 
a movement in triple measure, which they called minuet. 

But the peculiar grace and charm which Haydn, Boccherini, and 
Mozart infused into their minuets are lacking in those of Corelli, 
Couperin, and Bach. Hence the author classes the ideal minuet 
with modern dance movements, for it really is a product of the 
latter part of the eighteenth century, and belongs to music's most 
melodious epoch. The famous minuetto in Boccherini's A-major 
quartet may be cited as an example. The grace of this simple 
movement is such that conductors of symphony orchestras fre- 
quently include it in their popular programs among the numbers 
for string orchestra. The first section is here appended : 

Ex. 172. 

Nearly all modern minuets commence on the third beat which, 
therefore, receives either a rhythmic accent or a melodic punctua- 
tion so long as this peculiarity is manifest. (See Ex. 172.) 

The minuet was derived from the ancient peacock dance, and 
retains the ceremonial character of that obsolete movement. The 
performer should imagine a goodly number of ladies and cavaliers 
disposed in couples on the floor of a ball-room. The music begins. 
The dancers salute each other in graceful obeisance, and then pass 




in curving figures, each cavalier extending his elevated hand to the 
lady opposite, who offers hers in return, and thus they glide by. It 
was not so much a dance as a pantomimic promenade in which 
courtly grace, elegance of manner and dignity of carriage combined 
with rich costuming and brilliant surroundings in presenting a cap- 
tivating and harmonious picture. The movement of such a minuet 
{Ex. 172) is very moderate, and since grace and pose are the prin- 
cipal characteristics, the tempo should not be rigid, but slightly 
variable and yielding. Neither should the accents be strongly 
marked, for that style is inclined to suggest angularity of move- 
ment, as in the rigaudon and the czardas. 

There are minuets which begin upon the first instead of the third 
beat, for instance the one in Mozart's last E-flat symphony. Yet 
even here the third beat frequently comes into prominence in such 
places as these : 

And the favorite menuet ancien, by Paderewski, is another instance. 
The first period begins upon i, but the second and third periods 
begin unmistakably upon the third beat. So does the repetition of 

the initial period : EX. 174. 

"g~ ~r~f 9 . ~ r ~ 

The coda also starts with a preliminary note. Mr. Paderewski's 
minuet in G-minor manifests the same tendency in its rhythmic 
grouping. If the music does not divide itself naturally in this 
manner the student must not include punctuations nor special 
accents for the sake of an arbitrary formula. But in nearly all 
modern examples the peculiar features here mentioned will be 



found, upon close examination, to exist, and their presence must 
influence the interpretation. This statement has been called in 
question on account of the seeming exceptions, but the criticism is 
a superficial one. The author recently examined fifty minuets by 
standard composers, and found that forty-four began upon the third 
beat. Since then he is still more firmly established in his original 
belief (expressed in his " Musical Analysis "), because it is founded 
upon very substantial facts and circumstances. 

All such examples as the minuet from Haydn's " Oxford" sym- 
phony (Ex. 102), from Mozart's last G-minor symphony, and the 
minuet in von Weber's sonata, op. 24, require frequent punctua- 
tions after the second beat of certain measures, or a corresponding 
accent upon the third beat when it is the beginning of a rhythmic 
group, thus : 


Ex. 175. 

The emphasis on the tied note is principally on account of the 
syncopated character of these phrases. See also the minuets in 
Schubert's "Tragic " and B-flat symphonies. 

Polonaise. The modern polonaise is a Polish dance, and, like 
the minuet, ceremonial in character. There are several peculiar 
features to be observed in the performance. The rhythm of the 
theme is frequently syncopated, with the rule of accent reversed. 
The third beat is usually unaccented, though the second beat often 
comes into prominence. The Castanet rhythm of the bolero nearly 
always appears in the accompaniment, either in part I or part II. 
A few quotations are made from Beethoven : 

(or 42). 



At (a) the syncopated rhythm and mode of performance are 
shown. The accompaniment rhythm at (b) appears in the second 
period to enliven the movement. At (c) may be seen the peculiar 
cadence which always falls upon the weaker parts of the measure ; 
sometimes on 2, but more often on 3. Many of the groups begin 
upon the last eighth note, and this must be duly considered. 

Chopin, in his polonaises, varied the rhythm considerably, but 
preserved the characteristic features. In op. 40, II, the special 
accents fall mostly upon 2. 

The character of the modern polonaise is majestic and inclined 
to seriousness. These qualities exclude a quick tempo. Even the 
"military polonaise," which is marked allegro con brio, is taken at 
a moderate pace so far as the quarter note beats are concerned ; 
but the full harmony played in sixteenths gives the effect of an 
allegro movement. The average tempo is about ( I = 100). 

Bolero. This is another triple-beat movement, and as the casta- 

net rhythm, 



, is a prominent feature, 

one might be led to suppose that the bolero is similar to the polo- 
naise. There is, however, very little similarity. The bolero is a 
national dance in Spain, unlike any other terpsichorean movement, 
excepting, perhaps, the cachoucha. Originally, the bolero was an 


amorous dance or pantomime, accompanied by castanets to mark 
the mensural rhythm. A certain amount of levity and considerable 
animation characterize the best examples of bolero. The principal 
rhythm of the theme is so animating as to create the impression of 
a fast movement, though in reality the quarter-note beats seldom 

exceed ( J = 104). The rate of speed is, however, quite variable 

when the bolero is danced as as a pantomime. Another peculiarity 
of the bolero is its dual rhythms. Triplets of sixteenths, inter- 
spersed with other note values, and frequently combined against 
even notes in the accompaniment, produce a contentious, disturbing 
effect which is peculiar to nearly all Spanish dances. 

Moszkowski, in his op. 1 2, has not failed to incorporate all these 
characteristics. Chopin went still further in his op. 19. The pan- 
tomimic character of this work is unmistakable. The Spanish 
rhythm, the castanet accompaniment, and the weird, fantastic con- 
figurations of the dancer as he portrays the joy and despair of 
love, all are expressed in Chopin's music. 

The accents and the manner of phrasing can not, even in a 
general way, be prescribed. But the principal features have been 
pointed out, and these are intended to be of assistance in seeking 
the proper interpretation. 

Tarantella. This popular Neapolitan dance is one of the fast- 
est movements known. It is nearly always in "S- measure. The 

movement is presto or prestissimo. A few examples are in *I 


n n 
r f^f j 

with triplets of eighths, 

, which correspond to the 

regular whirl of six-eighths, ft 


has an uneven, leaping tendency, 

. The other rhythm 

this is represented by 

j J.2 



, which is less characteristic than 

the example in 3F. These two alternating rhythms are prevailing 

features of the modern tarantella. To these must be added the 



rapid, uninterrupted movement and the marked contrasts between 
soft passages, interrupted by sudden strokes, like the schlag of 
gong and cymbals in Berlioz's Roman Carnival. 

The tarantella is an old Italian dance which (like the polacca and 
minuet) has been fairly metamorphosed by modern composers into 
a new form. Originally, the tarantella was employed as an antidote 
to the sting of the tarantula, the old dance being a species of 
incantation. This singular characteristic of the old dance has sur- 
vived because of its appeal to the imagination of composers, and 
the best examples of tarantella express more or less faithfully the 
influence of superstitious association. The following excerpts are 
selected : 

Ex. 177. 

J. Rheinberger. Op. 13. 

- ru J>! r> PT-, ^L. 

^ g. *T I *T Ll ?J_l I ,1 |i 1 j I l_l ^m 

The sudden accents (sfz and fp) are more strongly contrasted 
here than they would be in a minuet, gavotte, or barcarolle. In 
fact, the special accents are often harsh and pulsatalic, and the 
characteristic effect depends upon an instantaneous change in 
dynamic quantity from loud to soft, or vice versa. The tarantella 
by Rossini (" La Danza," arranged as a piano duo by Liszt), 
Chopin's op. 43, Rubinstein's op. 6, all contain the characteristic 
features enumerated. Most of the phrases embrace four measures, 
but this rule must be verified by actual analysis. Also the par- 
ticular part of a measure where rhythmic groups begin must be 
definitely ascertained before a correct performance can be success- 
fully attempted. See the revised edition of S. Heller's taran- 
tella in A-flat, and that fine one by Rheinberger, composed as a 
piano duet, op. 13. 

Czardas. The principal characteristics of the czardas were 


known since the conquest of Pannonia by the Cythians, centuries 
before this dance was utilized in artistic composition by Haydn, 
Beethoven, Schubert, Liszt, and Brahms. 

The czardas is the most unique of all national dances. It has 
two parts, a slow and a quick movement, and the measure is 

A- O 

almost invariably W or *g-. 
TT * 
The first part, lassu, is funereal, almost tragic, in style. Part II, 

fries, is impetuous and wild in its expression. Sudden accelerando, 
ritenuto, or other deviation from the allegro movement mark this 
part of the dance. A variety of rhythmic figures are employed, 
though syncopation is the most common, particularly in the lassu, 

I"! "1 I 
J.J Jj m . The eminent Hungarian master, Mr. Franz Korbay, 

in his introduction to the original Songs of Hungaria, attributes 
this peculiarity of the music to the organic construction of the 
language, many words having their main accent on the second 
syllable. It is similar to an iambus, but with two accents instead 
of one. From this it would seem that the czardas is accompanied 
with singing, and so it frequently is. While rhythmic device is the 
most pronounced feature, the contrasts between loud and soft are 
characteristic, and the mixed modes are also peculiar. An example 
is quoted from a native composer : 

Ex. 178. 

Andante con Exp 

The symbols (to be strictly observed) were indicated by the 
composer. The lassu admits more expression than does the dance 
which follows, though even this is frequently ad lib. in style. 
In both movements the cadence is peculiar, and seldom ends upon 
the main accent : 




Ex. -179.' 

~8~~ ^"* 1 

These bear some resemblance to the close of the polonaise, and 
are what M. Lussy called " feminine cadences." The close of the 
lassu is retarded, but the close of the fries is a tempo, frequently 

There are two Czardas albums published for four hands, which 
will materially assist the student who may be interested in this 
peculiar form. The catalog numbers are : Peters, 1487, and 
Litolff, 997. Considerable care has been taken by the arranger, 
F. Behr, to indicate the style, expression, and nuances, and these 
may be considered fairly authoritative. After these have been 
mentally digested, selections may be made from the famous twenty- 
one Hungarian Dances arranged by J. Brahms. These are issued 
for two and four hands, violin and piano, and for full orchestra. In 
fact, this Hungarian form has been employed by so many modern 
composers that the student can not properly ignore it. The czar- 
das is frequently very irregular in its period construction, three- 
measure phrases being of common occurrence. And in truth the 
rhythmic effect is not always satisfactory to those who are not in 
full sympathy with Hungarian temperament and aspiration. 

Habanera. In Cuba the habanera is danced by a single person 
on a table or elevated platform. The habanera has two periods, 
which are many times repeated. It contains the peculiar features 

of Spanish dual rhythm, but is in ^, whereas nearly all the old 

Spanish dances are in 

, or j f measure. (Moszkowski's 



five Spanish Dances, op. 1 2, are in triple measure.) Arrieta, Mon- 
asterio, Bizet, Gottschalk, Neustedt, and others have helped to 
popularize the habanera in Europe and North America, and it must 
also be said that they have enlarged considerably upon the original 
form and character of the dance. A quotation is made from " Car- 
men " : 

Ex. 180. 



The movement is moderately fast. In addition to the uneven 
rhythmic combinations, it will be observed that some of the synco- 
pated effects are similar to those of the czardas, but not so strongly 

The habanera is essentially gay and animating, but rather impa- 
tient in its expression. Gottschalk's Cuban dances are good 
examples. See also the Avanera by M. Garcia, transcribed by 
J. de Zielinski. 

A discriminating study of these various species of dance form 
can not fail in emitting considerable light upon the exposition of 
Music's mysterious language. We need a tangible nucleus from 
which to proceed on our impalpable journey ; a guide which, like 
the mariner's compass, will at least indicate the general direction 
of our course. 

Whatever relates to measure, rhythm, movement, will be more 
easily acquired from the standard dances. 

In addition to this, and to the applying of special accents, a 
thorough understanding of the more important dances reveals 
many racial idiosyncrasies, which are expressed in larger and more 
artistic forms. 

It will be necessary hereinafter to refer to certain dances by way 
of illustration. 





Unfortunately, there is considerable diversity of opinion concern- 
ing the application of certain ornament signs. But in some in- 
stances it is perhaps better that all authors do not agree, because 
the absence of an established rule in regard to a certain ornament 
enables the performer to consult his own taste which is frequently 


Ex. 181, 

the truest arbiter in such matters. 

The mordente, or passing 
shake, is an example. This is 
always illustrated with the ac- 
cent on the first note of the pass- 
ing shake, whether written in 
small notes or indicated by the 
sign \v : 

Of course, there is a reason for this, otherwise it would not so 
universally have been translated in the manner shown. No doubt 
it imparts variety to the rhythm, and also accentuates the melodic 
tone thus ornamented. But the questions arise: (i) Does the 
melody admit this rhythmic alteration ? (2) Do the ornamental 
notes gain or lose by this accentuation ? 

In the author's opinion, this mode of performance (Ex. 181) 
applies more particularly to the old harpsichord and clavichord 
music. In modern works, especially since the advent of Chopin, 
the mordente is frequently to be considered as representing adven- 
titious grace notes, whose value is taken from the previous, not 
from the principal, note, thus : 




Slumber Song. Schytte. 


This causes no interference with the melody note nor with the 
rhythm, provided the grace notes are played immediately before 
the principal note, a, and without accent. Compare this with the 

customary mode of performance : 

Ex. 182 (a) is not intended to illustrate the prall triller, but rather 
those instances in which unaccented grace notes were intended. 
If the style seems to demand a transient shake, then the solution 
must be according to Ex. (b). But where unmeasured grace notes 
would be more appropriate, they should be performed before the 
time of the principal note and without accent, as in Ex. 182 (a). 
Esthetically, there is a wide difference between the two modes of 
performance, and therefore it ought not to be a matter of serious 
doubt as to which effect was intended in particular cases. The un- 
accented fioriture are much daintier and better suited to graceful or 
ethereal subjects. The Canzonetta by V. Hollander is an example. 
A fragment of this is quoted according to the printed copy (a), and 
in the more exact notation (b) : 

Ex. 183. 





The melodic outline is more nearly preserved at (b) than it would 
have been had the grace notes been interpreted as a prall triller. 
The lighter character of the grace notes, and the fact that they do 
not interfere with the principal tone, incline the author to favor the 
unaccented mode of performance in certain instances, especially 
since the prall triller is frequently angular and precipitant in effect. 
But in* the majority of instances the customary method is to be 

An effect similar to that of the prall triller is sometimes indi- 
cated by tr., particularly in quick movements : 


Ex. 184. 


Owing to the rapid movement it is not practicable to undertake 
a regular trill at the places marked tr. Therefore the passing shake 
{which here sounds like a prepared grupetto) is sufficient. 

In these instances the signs tr. and \v are identical in their 
application. Thus, from a Preludio by Handel : 




-TT I Zj 

In a foot-note by von Bulow the tr. is translated in this manner,, 
precisely as though the sign \V were employed. But under ordi- 
nary conditions they are treated differently. There are, therefore, 
three different methods (a, b, and c) for indicating a single effect, 

= S= (c). 

Ex. 186. 


In a fast movement the tr. is 
played more like a triplet figure, 
which serves to distinguish it from 
the mordente : 

Taste must decide between these distinctions, for in Ex. 185 (a), 
the tr. is translated as an inverted mordente (b). 

The choice between a couplet of grace notes and a regular prall 
triller must also be referred to the esthetic sense : 

liar ghetto. 


The difference in effect is, however, so great as to leave no room 
for doubt in such instances. The graces at (a), written in large 
notes as an illustration, are light and graceful ; at (b) the prall 
triller sounds rather angular because the ornamental notes fall upon 
the main accent. In the second arrangement the rhythm of the 
melody is almost destroyed and the slurred staccato style is inter- 
fered with. At (a) these features are preserved. 

But if the \v were reserved to indicate the prall triller (Ex. 186, 
d), and the small grace notes were used only to express the unac- 
cented fioriture (Ex. 183, b), all doubt would be removed. 

The abbreviation tr. usually carries with it something more than 
is indicated by the sign, iv, and here is another cause for confusion, 
since their application is sometimes identical.* 

Where the conclusion is indicated, as in Ex. 184, the triplet 
figure is a sufficient expression of the abbreviated trill. Where no' 
conclusion appears, the performer would ordinarily play five notes 
for the tr. The second period in Beethoven's E-flat minuet pre- 
sents an example : 

Op. 31, III. 

Ex. 189. 





-J K L 

* It is useless to hope for a more precise mode of representing these ornaments, be- 
cause the task of writing notes is usually so irksome to composers that they will natu- 
rally avail themselves of every sign or symbol which serves to abbreviate their work. 




The original notation appears at (a). The arrangement at (b) is 
vocal in character and has this advantage : it does not anticipate 
the following melodic note, g. In view of the lyric character of 
this minuet the translation at (b) is perfectly euphonious and 
proper. At (c) a passing shake in grupetto form is given. At (d) 
the trill is so written as to leave the important melodic note, g y for 
the following measure. And whenever the trilled tone passes 
below, instead of above, this plan may be adopted ; for there is no 
good reason why every trill should have a concluding note below 
the trilled note. See " Anitra's Dance," by Grieg. 

Of course, the pianist may include seven notes (in place of five) 
for the trill, if he so elects. All these considerations must be 
weighed in choosing between different methods of performing em- 
bellishments, unless the further circumstance of epoch* intervenes. 

The grupetto now claims attention. Since there are various 
methods of performing this ornament, style and movement should 
have a determining influence. Suppose this example occurs : 

Ex. 190. 

A turn is indicated by the symbol, but we know not which of 
the following readings to adopt until the character of the music is 
duly considered : * 

* This will be discussed in chapter xxx. 



If the style were distinctly "lyric, and inclined to seriousness, (a) 
would be preferable. At (b) the absence of accent gives more 
prominence to the principal melody note and less to the grupetto, 
because the quintole here is to be played equally, not as 3 and 2, 
nor as 2 and 3. The third illustration (c) is applicable to a livelier 
or more energetic strain. 

Where the symbol is placed directly above (or below) the prin- 
cipal note, it is usually better to begin with the latter, in order to 
give due prominence to the melody note. Paderewski's Minuet 
a la Mozart illustrates this : 

Ex. 192. 

The only difficulty consists in equalizing the group of five notes. 
Performers frequently give an uneven or disjointed effect to these 
groups by dividing them into two parts, thus : 



The most remarkable example, and one that applies directly to 
our studies in nuance, is " Eusebius," from Schumann's Carnival. 
The entire movement is evolved from ad libitum grupetto figures : 

Ex.193, (a). 


Op. 9, Y. 

Sotto voce. 



The major portion of this period is performed as if it were in - 


measure, and though the composer's notation is not a very exact 
representation, it indicates the main divisions where the pressure 
accents occur : 

In the second period the grupetto appears as a quintole, followed 
by a triplet. These also are to be equalized, but in a peculiar, 
tempo rubato manner, which no words nor symbols can express. 
A discriminating study of this number from Schumann would prove 
of great assistance to the inexperienced performer in helping him 
to a better understanding of the turn, and all similar ornaments. 
No other embellishment is so frequently misconceived as is the 
grupetto. The general tendency is toward either overaccentuation 
or want of smoothness. Whereas the turn is usually an orna- 
mental group of unaccented notes woven around a principal melody 
note. In lyric themes the turn should be executed with extreme 
smoothness (by means of the legatissimo touch), and with very 
little accent. Slow movements naturally admit of a more repose- 
ful, expressive delivery of the grupetto, mordente, appoggiatura,. 
and other embellishments. Here more time is allowed for the per- 
fecting of the nuance, and also there is a greater demand for 
the expressing of sentiments which must be clothed in graceful 

There is another style of ornament, sometimes called the slur,. 
which is to be noted here. It has a peculiar portamento or glid- 
ing effect when its tones serve as reinforcements of the two preced- 
ing melody tones, as here : 

Ex. 1 94. ( a ). Rhapsody II, in F #. Liszt. 



This occurs at the end of a strain marked ritenuto, where the 
.style is ad lib., and the nuance need not be sacrificed to the de- 
mands of strict movement. The glide (a) falls upon the first of the 
measure, but the principal accent should be reserved for the me- 
lodic tone to which the interpolated glissando naturally leads : 

Observe the dynamic symbol below. Also, it should be stated 
that the manner of representing the glide is not very exact, because 
it is not so rapid as the thirty-second notes would seem to indicate. 
A quasi portamento effect may properly be introduced here. As a 
contrast to the glide (and for reasons already stated) the inverted 
mordente in the third measure of Ex. 194 should be executed in 
advance of the beat, and without accent. 

An interesting illustration is taken from Beethoven's seventh 
symphony : 

Ex. 195. 

VIOLIN II. Allegretto. 


This charming counter-subject occurs beneath the monotone 
motive of the second violins. The glide comes upon the first of a 
measure, yet without altogether displacing the accented principal 
tone. The symbol =dl is therefore to be interpreted literally. 
A peculiar pressure touch is required to give the proper effect, 
which belongs primarily to violin and vocal music. The glide 


here is not so rapid as is the passing shake, yet the principal tone 
and the rhythm of the measure must be preserved as nearly as possi- 
ble. An ad lib. style of performance is required, but this must not 
interfere with the regular, measured movement of the accompani- 
ment. The precise effect, like certain foreign idioms, is difficult of 
translation. The arrangement (a) is too angular and destroys the 
grace of the ornamentation : 

Ex. 196. 


* ~*~ ~ w ~ m 

The following (b) is more nearly correct, though even this is in- 
clined to sound too trivial and sprightly. 

The nuance is worthy of serious consideration, and will require 
both practice and discrimination. Even then no pianist can rival 
the nameless charm which a fine body of string instruments impart 
to this romantic sub-theme. 

As a resume of this chapter, the following instances are recom- 
mended for practice : Serenade by P. Douillet, op. 6 [M] . Ber- 
ceuse, op. 23, VII, L. Schytte [M]. Neither the appoggiaturas 
nor the fioriture should be played too quickly. Rondo in E-flat, 
von Weber [M D]. All the graces are written in large notes, 
but without accent : * 


Ex. 197. 

Valse Caprice, op. 16, KarganofT [M D]. Gavotte in rondo 

* These graces are miscalled by Germer and other writers " double appoggiaturas." 


form (F-major\ Padre Martini [E]. The passing shakes are ac- 
cented. Sonatina, Beethoven, op. 49, I, first movement [E]. 
Finally, the Nocturnes and Ballades, by Chopin. 

Also the " Embellishments of Music," by Louis Arthur Russell, 
is particularly recommended. 


The author's primary definition of rhythm refers to the order or 
arrangement of notes in a measure with regard to their value. For 

instance, the castanet rhythm of a bolero : 

This is a prominent feature of the dance, and continues almost 
incessantly from beginning to ending. In like manner the com- 
poser may suggest by means of fundamental rhythms the barca- 
rolle, the lullaby, or the spinning song, thus : 

Ex. 198. 




Of course, this rhythm is dependent upon measure and move- 
ment, but may be considered quite independent of melodic phrase 
or harmonic relation. Rhythm is the mainspring which gives to 
music its forms of motion and repose, turbulence and quiet. It is 
the motive force without which the spirit of a composition becomes 
inert. Hence the characteristic features of certain dance rhythms 
as indicating particular forms or kinds of motion. 

In this limited, though important, sense are to be considered 
various syncopations, fictitious note -values, combination rhythms, 
and other arrangements. 

The larger definition of this element includes the dual rhythm of 

a phrase, 

; also a series of rhythmic groups con- 

sisting of several measures : 


Ex. 199. 

RHYTHM. 153 

Barcarolle. Jensen. 

The entire passage consists of a rhythmo-melodic pattern or 
figure, and this is repeated in sequence form until a different pat- 
tern appears. The arrangement of notes with regard to their time- 
value is the same throughout, and therefore the entire cadenza is 
comprehended as one extended rhythm. The composer joined this 
together by means of the continuous slur. 

Occasionally a variety of figures are included within a rhythmic 
phrase and slurred together. In the following excerpt the phrase 
points (indicated by interrupted slurs) are equally distributed : 

Ex. 200. 

W. H. Sherwood. Op. 5, II. 

These rhythms include regular phrases, which are to be punc- 

In this larger sense, rhythm is produced by the regular recur- 
rence of orderly groups, embracing principal and subordinate 
accents. (Exs. 199 and 200.) This, however, soon becomes 
monotonous unless relieved by a change in measure, form, or 
movement. Perhaps this is the simplest explanation of syncopa- 



tion and other mensural or rhythmic deviations, though frequently 
they serve a higher purpose than that of adding variety to the 

An instance is quoted from Chopin's D-flat waltz, op. 64, I : 

Ex. 201.. 

The four melody notes played against three quarters in the base 
produces a syncopated effect, yet much more consonant to the 
theme than either of these arrangements : 




No fictitious note -values are here represented, but the rubato 
effect of the original is sacrificed. Something similar occurs in 
Schumann's op. 124: 

- """" Schlummerlied. 

Ex. 202. 

2 <*~ I- *>. I 

i2: ^1= iMnqifcri -dzi 

The manner of performance is here indicated by the figure 2 
placed over each melodic division of the first measure. The effect 

is that of 5 against 3f By dividing this into \p- it will appear 

that the actual value of the notes, is not altered, and yet the effect 
of dual measure is produced : 



The common arrangement of the melody would be, 


in comparison with which observe how much more earnest and 
impatient is the expression according to Schumann's notation ! 

This composer was a master of rhythmic device, and though 
certain vagaries have been charged against him, his scores reveal 
many instructive and effective examples of unusual rhythmic ar- 
rangement and grouping. Indeed, some of these which have 
incurred the displeasure of pedagogs are not more peculiar than 
are the extended thematic melodies of Chopin or the Norse har- 
monies of Grieg. 

Music in the rubato style frequently requires such alterations as 
are shown in examples 201 (a) and 202. But to change a measure 


from 1? to 

produces such a radically different effect that no 

one should take such liberty without full assurance as to the artis- 
tic result. It has, however, an advantage over the regular divi- 
sions, especially if the accompaniment is in notes of the same appar- 
ent value, thus : 

Ex. 203. 

In addition to the esthetic effect of 2 against 3 (as pointed out 

j 5 6 


in the last Schumann excerpt), this dual rhythm brings the theme 
into greater prominence because the second and fourth melody 
notes do not fall simultaneously with the accompaniment, but are 
heard independently. Another peculiarity is to be noted : any 
combination of even and uneven rhythms (such as 2 against 3, or 
3 against 4) expresses something of contention or impatience, and 
these qualities of dual rhythm must be duly considered by the per- 
former, as well as by the composer, whenever they are introduced. 
Unusual, and even complicated, rhythmic designs have entered 
so freely into the manifold expression of modern music that the 
performer must boldly face the problem and endeavor to solve it. 
He can no longer evade the question, as Christiani did, by attribut- 
ing these dual rhythms to the vagaries of a distempered brain, 
nor even to the trivial idiosyncrasies of composers. The technical 
material of the art has been greatly enlarged during the present 
century, and the scope of musical expression has been extended to 
a corresponding degree. The simultaneous use of two kinds of 
measure may have been the composer's best method for represent- 
ing a dual state or of bringing forward an opposing element. The 
waltz in A-flat by Chopin, op. 42, is a farniliar example. Let no 
one suppose that the composer intended such effects as these : 

I i 

or ** 

In the original a melodic accent falls upon the notes with upturned 
stems, but this does not necessarily destroy the triple measure, 

since the left hand plays its rapid waltz rhythm, j | | I I 
in regular tempo. Compare the following arrangements, in order 
more particularly to note the syncopated effect at (b), which is the 
-original : 



The latter is much more confident in expression, owing to the 
inciting, impelling character of the syncopated anticipations above. 
It represents one of Chopin's brightest moods, lightsome and gay, 
but impatient of restraint. 

The following example, from Nicode's Ball Scenes, is more 
easily managed : 

Ex. 205. 



/ cres 


LJ ! U I Ld 

f^-# s! *^9-9 ^=*- r * ^0-9 i 

It was not here necessary for the composer to alter the mensural 
signature, since the accompaniment above continues in waltz meas- 
ure. But the performer's sense of rhythm must incline him to 
interpret the passage in this manner : 

because it is the only solution of the rhythmic anomaly. This is 
not a tempo rubato, but a rhythmic effect. Observe that the con- 



tents of a measure at (b) do not equal a full measure at (a) ; in 
other words, the time-value of the notes remains the same. 
It is different here : 


Ex. 206. 

-9 -t-9-9 

This is a combination of the " Nibelung " and the " Siegfried" 
motives as opposed to each other. The latter is an antithesis to 

" is therefore 

the former, and the combination of against 



significant and appropriate. The eighths above are, of course, not 
so quick as those below. 

A noteworthy instance of unusual rhythmic division occurs in the 
finale to von Weber's Concertstiick : 

Op. 79. ( From Liszt's Edition. ) 


The twelve sixteenth notes in each measure are here divided into 
three equal rhythmic groups, and in order to remove all doubt as 
to his intention, the composer further directs that each group shall 
be accented. This is considered by certain pianists to be unnatural 
and whimsical, but quite the contrary is true. The melodic con- 
tour falls naturally into groups of four notes, 

RHYTHM. 159 

and these admit of rhythmic accent as freely as do other forms of 
syncopation. The utmost precision and exactness must, however, 
be observed in the performance, for the unusual grouping does not 
in any manner affect the actual note-values. There is, therefore, 

no reason why the * measure may not be maintained in the ac- 

companiment, though the passage bears considerable resemblance 

to *F measure. An example somewhat similar may be found in 

the Divertisement by N. Stcherbatcheff, op. 8, IX. The twelve 

sixteenth notes (-JF measure) are divided into four triplet groups. 

The careless manner in which certain composers write ^^ over 

two triplet figures causes frequent misapprehension, and may be 
adverted to here. In strict designation a sextolet is almost 
opposite in its effect to that of two triplets. The same difference 

o / 

exists between * and T measure : 

Ex. 208. 

A - - A > 

The first six eighth notes correspond to the effect of a sextolet ; 
the groups of eighths at (b) have an accent upon each division of 
three notes, and resemble triplets. Not alone is the rhythm altered 
by these different mensural divisions, but the melodic and harmonic 
characters also undergo change. 

If we write two sixteenth notes against each eighth note at (b), a 
sextolet will result : 

Ex. 209. 



A slight accent falls upon every other tone of the lower groups,, 
and this is what distinguishes a sextolet from two triplets. 

In the first andante of Rossini's " William Tell " overture there 
is a group of two triplets marked as a sextolet : 

What makes this still more misleading is that the six notes are 
joined together by a single bar, as though a sextolet was actually 
intended. Of course, an experienced player will recognize at 
once that two triplets were contemplated, thus : 


The "^ is merely the result of carelessness. A similar mistake 
occurs in the " Zampa " overture : 

Ex. 211. 

The so-called sextolet evidently embraces two triplets, and should 
have been so written. The cadence in Kirnberger's C-sharp minor 
fugue presents another instance. 

Even Grieg, who is usually very particular about transcribing 
his thoughts, commits the same error in his op. 19, II. See first, 
period, groups marked ^^. These should be ^^^^. 



The reverse of this rarely occurs, though an instance appears in 
the largo of Beethoven's op. 7 : 

Ex. 212. 

: /"3-\~*~s: 1-HJ8-f-fu-3g : - ->- ^ I ps-f 0-f- V-^-SH I 



This is from the Stuttgart edition, and we know not whether the 
last group of six thirty-second notes was indicated by the master 
or by Herr Lebert Without regard to this matter of responsi- 
bility, the author believes the last group should have been denoted 
as a sextolet, which is rhythmically more correct. Perhaps it will be 
said that the six thirty-second notes are the same melodically as the 
two triplets at (a). The last measure of our quotation, however, 
is not a repetition of the first measure, but a continuation the 
sequential group at (b) being a third higher than it is at (a). The 
natural thematic development in a simpler form would appear like 

Ex. 213. 



When at the close the number of 
notes is doubled, they naturally fall into 
the order of a sextolet two thirty- 
second notes against each sixteenth note 
of the triplet, thus : 

This corresponds also to the first appearance of this intermezzo 
where the four sixteenth notes of the first phrase are answered by 
eight thirty-second notes in the second phrase. Therefore the 
sextolet (Ex. c) seems to the author more correct and effective, 
especially from a rhythmic view-point, 



For similar reasons the sextolet in the rondo of this same sonata 
is correct : 

Ex. 214. 

An interesting example occurs in the scherzando of Saint-Saens, 

op. 31. The groups of- six sixteenth notes (see Ex. 198), which 

belong naturally to part I in |T are maintained in the accompani- 


ment after the melody changes to ^3-. The groups of six six- 


teenth notes are then marked as sextolets in order to preserve the 
spinning-wheel rhythm by means of uniformity in the accompani- 

Other writers have evidently had experience with these ambiguous 
notation signs. Witness the accent marks which Mr. Sternberg 
placed on every alternate note of the sextolet quoted in Ex. 53, 
chapter vi. 


Capriccio, B. Stavenhagen, op. 5, I. Gigue, B. O. Klein. 
" Biroulki," A. Liadow, op. 2, IX and XIII. Ignace Brull, " In the 
Mill," op. 72, VIII ; Rondo, op. 71, IV. 


Prescribed Movements. While it is true that an absolute, un- 
varying movement is contrary to the spirit of music, there are cer- 
tain forms which may be said to have a prescribed rate of speed. 
The march, the galop, and the waltz are common illustrations, 
though the latter sometimes contains ritenuto passages. The 
various species of dance form, treated in chapters xiv and xv 
may be very nearly indicated in their movements by means of 
metronomic symbols. - The czardas is perhaps the only exception, 
though epoch may have a determining influence in this matter. 

When Beethoven marked his op. 42 " alia polacca" he clearly 
indicated that it was to be performed in the style and according to 
the general movement of a polonaise, but with more freedom than 
is allowed in the ball-room polonaise. Also, tempo di menuetto 
indicates the movement of a minuet, but allows some liberty of 
style and nuance. Therefore w r hen a composer selects a dance 
form in which to embody his mood or expression, the peculiar fea- 
tures of this particular style are to be suggested, even though the 
dance be purely ideal in its character. But in these instances 
movement signifies something more important than the rate of speed 
at which the music moves. It includes that peculiar/<?r;^ or kind 
of motion which is suggested by the rhythm of any characteristic 
dance. The minuet and the polonaise are good illustrations. The 
form of accompaniment, the peculiar trend of the melody, a bit 
of syncopation, or a delayed cadence, all these features enter into 
movement and impart to it that suggestion of individual, charac- 
teristic motion which we recognize in a good polonaise or minuet. 
All this is chiefly influenced by rhythm. 

Many of the miscellaneous single forms might be classed here. 
Thus the motion of a boat upon the water is approximated by 



means of a certain rocking rhythmic arrangement included in a 
regular mensural movement of beats. Also the cradle spng and 
the spinning song each have a prescribed movement, though more 
freedom of style and nuance are allowable in these instances than 
in the dance form. 

Variable Movements. The rhapsody, the fantasia, and the 
caprice illustrate one phase of this subject. 

Moods which are alternately grave and gay, like shadow and 
sunshine, necessarily require a changing movement. The Hun- 
garian Rhapsody in G -minor t by Heinrich Hofmann is a pro- 
nounced type of this kind. First there is an animating allegro, 
followed by a cantabile which requires a more quiet movement. 
Then the first allegro recurs. This is succeeded by an intermezzo 
which ends/0f0ttY., in anticipation of the piu lento. Here are two 
phrases reminiscent of the lassu, interrupted by four measures of 
impatient allegro. In this capricious manner the music continues, 
each different period being at least slightly changed in movement. 
Instances of this kind are so numerous that it is unnecessary to 
enumerate them. 

Contrasting subjects, such as " Pierrot et Pieret," " Pantalon et 
Columbine," naturally demand a variable movement ; and these 
opposing themes frequently occur in larger and more serious 
works. It will prove more instructive if we select as an example 
the first, and usually most important, allegro movement from a 
sonata. The lyric theme, whether first subject or second subject, 
is not quite so quick as is the thematic subject. When these two 
styles are included it is for the purpose of contrast, and a lyric 
theme will present but little contrast to a thematic one if the same 
movement be maintained. Several instances are cited from Bee- 
thoven. In the first movement of op. 2, I, the principal theme is 

( I = 112); the second theme, ( I = 120); the conclusion, 

& & 

( \ = 104). There is not much contrast between the first and 
second subjects, one being an inverted form of the other ; but the 
second subject is in a mixed mode, and more animating. The con- 
clusion is ad lib. in style. 

The opening allegro of op. 7 begins brilliantly, and is mostly in 


thematic style. The second theme is a lyric, and here the move- 
ment is considerably lessened. Then a stringendo passage leads 
back to the allegro con brio. The final close is a stretto, and there- 
fore to be played accelerando. 

See also the Andante and variations, op. 26. The movement 

varies from ( r = 72) to ( r = = 96), and in the last variation 

the recollection is played about { r = 69). 

The A-minor waltz by Chopin, op. 34, II, is an interesting study 
in connection with this subject. Each of the four separate periods 
is taken at a different pace. The first is slowest ; the second is un 
poco vivo and quite variable ; the third is in regular waltz tempo : 

E, 21 ,E^ 

The fourth period, sostenuto, is slightly less animating, and its 
counterpart in A-minor (IV, b) is slower and more serious. The 
coda (period V) is faster than the initial period which precedes and 
follows. The final cadence requires considerable ad lib. nuance, 
and the last chord must be briefly sustained. Otherwise the end- 
ing is inclined to sound abrupt and unsatisfactory in comparison 
with the melodic beauty of the work. 

The classic overtures present further illustrations of altered move- 
ment. Overtures in the sonata form (such as " Figaro " and " Ruy 
Bias ") are not so variable in this respect as the more modern pic- 
torial overtures, like " Sakuntala " and Tschaikowsky's "Romeo 
and Juliette" or " 1812." 

Connected Movements. When the measure is altered in the 
midst of a movement, the usual intention is to preserve uniformity, 
either in the beats or in the measures. In Rubinstein's tarantella, 

op. 6, there are twenty-four measures in ^, and then the princi- 


pal strain is marked (as tarantellas usually are) St. Meanwhile, 

1 66 


the movement remains the same : i. e. t a measure of 5 occupies 


the same amount of time as does a measure of 7. The com- 

poser's object in using ^_ measure for the introduction was to 

represent groups of four sixteenth notes without resorting to ficti- 
tious note-values : 

Ex. 216. 

-^ 1 t 

After the second subject this recurs in form of an intermezzo, 

and in the coda there are four measures of ^~ > Du ^ from begin- 


ning to ending the movement remains the same, excepting the rallen- 
tando and accelerando passages. 

Mendelssohn's " Scotch Symphony" contains several interesting 

ja. . 
examples of connected movments. The andante, in -*-, is marked 

(g! = 72), and the principal allegro, in W is ( J . = 100). But at 

the close of the andante there is a pause, and after that the quick- 
ened movement enters naturally. At the end of the allegro there 
is a descending unison passage leading to a fragment of the an- 
dante, which closes the movement. The latter is again indicated, 
( ] 72). But if the strict movement of the allegro be maintained 

during the connection of the "& with the 


, the want of uni- 

formity between ( ! = 100) and ( ! = 72) will produce an un- 
pleasant effect of discrepancy in the beats. This is to be avoided 


by means of a gradual rallentando during the last eight measures of 
the allegro, so that the movement of the andante will be approxi- 
mated before the measure is changed to ^f-. The last eight 

measures of the allegro were plainly intended to connect with the 
andante, and no incongruity should be perceptible in such con- 
tinuous passages. The measures here do not correspond, one with 

the other ; it is the beats ( i and J) which should be made uniform 

/ o 

in changing from 1 *- back to *. 

Another illustration is Beethoven's op. 13. The eighth notes of 
the grave are played very nearly in the same space of time as are 
the half notes of the following allegro ; and one measure of the 
adagio is equal to two measures of the rondo, approximately. 

Where the movements are separated by pauses there is less 
necessity for preserving uniformity in the beats or corresponding 
rates of movement. The greater number of cyclical forms (as 
sonatas and symphonies) should, however, be carefully considered 
with regard to the uniformity or correspondence of beats in the 
various movements, because this is one of the conditions of unity 
and continuity. Stavenhagen's interpretation of the op. 27, II, by 
Beethoven, gains immensely through the artist's peculiar method 
of catenation. For obvious reasons orchestra directors usually 

take a broader view of this matter than do soloists. 

In " Die Meistersinger," act II, the change in measure from ] 


to ^ was marked by Wagner ( i = ), clearly indicating that the 

beats retain their uniformity. Several similar instances occur in 
" Siegfried," and in the other music-dramas. 

In the termination to Schubert's " Rosamunde " overture the 

J . 3 r is equal to the I of the alia breve, and thus the accented 

beats correspond. Numerous similar examples might be quoted. 
The most exact precision is often required where a new measure is 
thus introduced uninterruptedly, and where groups of six sixteenth 
notes must be executed in the same space of time that was con- 


sumed by four eighth notes. Even the slightest discrepancy be- 
tween the beats in such instances would prove unsatisfactory, if not 

When the movement is changed in passing from the lyric to the 
thematic style, or vice versa, it is generally advisable to increase or 
decrease the tempo gradually, as explained in connection with the 
"Scotch Symphony." Thus in the first allegro of the " Wald- 
stein Sonata" the principal subject is about ( ! = 168), and the 

second subject is considerably slower. During the four measures 
beginning at 3 1 the movement should be retarded so that the lyric 
theme in E will enter naturally and without seeming to hesitate. 
Then from measure 49 a gradual crescendo and accelerando will 
restore the original movement ( | = 168). 

Tonal Considerations. The relationship of keys must also be 
considered in determining whether two movements are to be sepa- 
rated or played attacca. Mendelssohn's symphony op. 56 was 
intended to be performed continuously. The keys are either iden- 
tical or parallel : A-minor y F-mafor, A-minor, A-major. The tonal 
arrangement is similar in Beethoven's fifth symphony, which re- 
quires no pauses, excepting a brief one after the first allegro. In 
both works the affinity of motives is closely maintained, and this 
fact also is to be taken into account. In Beethoven's sonata in 
E-flat, op. 7, the largo is in C, and there is no apparent connection 
between this and the other movements. Therefore pauses should 
be observed after each movement, and all similar instances may be 
treated in the same manner. In the so-called " moonlight sonata " 
the key -tone 1 ' remains C^L (or D-flai), and this fact is, in itself, 
a sign of connection. Mozart was the first to write a symphony 
(the little one in D) y in three connected movements. Mozart pos- 
sessed, however, a wonderful gift of spontaneity, and while nearly 
all his cyclical works may be played continuously, we could not 
treat the average sonata in the same manner. 

Liszt's E-flat concerto and his symphonic poems are so written 
as to leave no room for doubt about their continuity. 

Influence of Rhythm upon Movement. There was at one 
time considerable misapprehension of Beethoven's intention in 



designating the second movement of his eighth symphony, " Alle- 
gretto scherzando " ; and the movement, measured by quarters 
instead of eighths, was usually taken much faster than Beethoven 
intended. While it is true that the quarter-note beats succeed 

one another rather slowly (about J = 48), the fact must be 
considered that the melody notes are mostly sixteenths and 
thirty -seconds : 

Ex. 217. 

An allegretto movement measured according to customary 
standards would be altogether too fast for the grace and humor of 
this number. But the term, allegretto, was intended to be under- 
stood in its literal sense, indicating a cheerful style ; and the short 

notes ( rT2 ^53) are not parenthetical or adventitious, but form 

* *3 *3 

part of a principal theme, and are therefore played prominently as 
melody notes. Compare these thirty-second notes with those in 
the introduction to the master's second symphony, or with the 
groups of eight thirty-second notes in his op. 2, I, adagio move- 
ment. These latter are unaccented, excepting when they fall upon 
regular metrical divisions. The composer was therefore justifiable 
in his use of the word, allegretto, though he seems to have pre- 
sumed that we would " read between the lines." The intention 

would, however, have been more plainly indicated by means of Jg 


in place of ^. 

Therefore rhythm has a governing influence in all kinds of music. 
The rondo brillante by von Weber is recognized among the lively 
movements ; yet the same quarter-note beats might represent an 
andante if the melody notes were of much longer duration. (This 
is explained more elementarily in chapter V of " Complete Musical 
Analysis," and need not be dwelt upon here.) 

In conclusion it must be stated that after certain deviations from 

i ;o 


a fixed movement the performer should be particular to resume the 

original pace when indicated by the term a tempo. 



Rhapsodic Guerrere, C. Binding, op. 34, VI (Peters' Ed. 2867, 
b.) ; Campane a Festa (Epitalamio), G. Sgambati, op. 12, VIII; 
Une Vision d'amour, Wrangell, op. 13, I; Lamentation, Ad. M, 
Foerster, op. 37, I ; Reverie Pastorale, B. Godard. 


In rhyming poetry we find a regular metrical arrangement of 
syllables and a certain rhythmic cadence. But in prose the metri- 
cal order is more free and less euphonious. A short and a long 
sentence may succeed each other, or there may be a paragraph 
containing simple and compound clauses. Compare the metrical 
arrangement in Gray's Elegy with an oration by Demosthenes or 
Cicero. A parallel distinction exists between the lyric and the- 
matic styles in music. 

Phrasing is " a method of expression by phrases " ; but if a given 
music piece (such as the first etude in C by Cramer) is not con- 
structed by means of phrases, we must not apply to it the princi- 
ples which govern the performance of regular periodized composi- 

Thematic music is motivized, and usually consists of scale work, 
arpeggio or broken chords, free sequence, passage, transition, or 
canonic imitation. It is the antithesis to lyric music. Compare 
Bach's Inventions with Mendelssohn's Songs Without Words, or 
Rheinberger's G-minor fugue with Henselt's Love Song. It is 
to be understood that the term thematic is applied only to music 
in which regular, tuneful melody is not a predominant feature. But 
in strict designation melody is never entirely absent from thematic 
or harmonic works. 

There was a thematic epoch in music extending from the time of 
Frescobaldi to that of P. E. Bach ; and prior to the advent of Mon- 
teverdi even the vocal works were thematic. 

In modern compositions it is not always easy to distinguish be- 
tween lyric and thematic designs. The fifth variation in Bee- 
thoven's op. 26, for instance, presents no external resemblance to 
the lyric style. Yet this variation is a slightly varied transcript 




of the principal theme. To apprehend this fact requires some 
familiarity with the original melody and its harmonic basis. All 
such designs hold in solution, as it were, a certain melodic sub- 
stance which the composer purposely screened from cursory view. 
Generally speaking, thematic music belongs to the intellectual 
phase of composition, and it is, therefore, more susceptible to defi- 
nite analysis. Mensural and rhythmic accents, and punctuations 
discriminately applied, are indispensable to the proper performance 
of thematic designs, because here we have no well-defined melody 
dividing itself naturally into phrases or sections. The meaning is 
more or less obscured by a multiplicity of notes, and the design 
seems intangible until it is successfully analyzed. For example a 
prelude in G by Haendel is selected. In Litolff's edition this 
appears without a sign of accent or punctuation, and if performed 
in this manner it would have no significance whatever : 

Ex. 218. 

Suite XIV. 

The entire prelude is evolved from this short motive : 

Every repetition or imitation of this figure is to be considered as 
an independent group or motive, and punctuated accordingly. 
There are two methods of indicating the design, and these are pre- 
sented : 



At (a) the last tone at the end of each slur is to be as short as 
possible (if the accentuation is omitted) in order to separate it from 
the following sequence figure. This is punctuation ; and since the 
groups are sufficiently distinguished by means of staccato, the spe- 
cial emphasis on the first of each figure becomes superfluous. At 
(b) the punctuations are omitted, and special accents upon the 
beginnings of groups are substituted. Each method should be 
practiced separately. 

The cadence is effected so naturally and the period is so regular 
that there is little probability of failure in making the design mani- 
fest. But the second period will require greater care. At first 
there are two rhythmic groups, each containing eight sixteenth 
notes. These should be punctuated according to Ex. 219 (a). 
Then there is a rhythm of sixteenth notes ending on d-sharp stac- 
cato. The same principles apply to the next phrase, which makes 
a cadence in E-minor. 

What follows is a motivization of the short subject quoted in Ex. 
218 (b). Every sequence figure is to be punctuated in the manner 
already illustrated, and a gradual crescendo should accompany the 
modulations to the highest point, a. 

The first descending group consists of twelve notes, the second 
of twenty, and each of these must be separated, or otherwise dis- 
tinguished. Then the original motive reappears. This and the 
preceding phrase are quoted : 

Ex. 220. 

|C}i-j| 1 1_. 1 |_|f * ~^ 

1 -r^-r : 

F '...: =b 

-t + i 

* ^ 


While the slightest ritard continues until the base makes its 
cadence on G, the re-entry of the main subject above should be 
sufficiently animating to attract attention, and to inform the listener 
that the original motive has, after a considerable digression, re- 
turned. Hence the accent marks at (a). A slight rallentando 
accompanies the last measure as far as (a) in order to emphasize 
the ending.* (This prelude may be found in LitolfT's Album 
Classique, No. 391.) 

Bach's two-part Inventions are now selected for analysis. No. I 
is constructed from a short subject which divides itself into two 
unequal semiphrases. The first is more important, as it constitutes 

the motive : 

The second half of the subject presents a contrasting rhythm, 
and serves a purpose somewhat 
similar to counter-subject in fugue : 

This is slightly less prominent than the motive (a). 

The method of grouping illustrated by examples (a) and (b) may 
be applied very generally to this invention. Thus the second 
phrase : 

Ex. 222. 

Under ordinary circumstances it is unnecessary to emphasize the 
first tone in these motive groups that is, when the preceding tone 
is detached. (See three last examples.) 

In measures 3 and 4 the motive appears inversely. This merely 
presents a different phase of the same subject, and is to be treated 
as principal theme. 

With regard to the reading of the following excerpt a difference 

* The pause is intended to be very brief. 



of opinion might reasonably exist. Several methods are pre- 
sented : 

fcq^-g-f-f-f i^r-h-i-F-^ 

The interpretation at (a) is in accordance with the original 
motive. This might be urged against it that the punctuating of 
every semiphrase is too literal an application of motive-grouping, 
and if continuously maintained, might seem pedantic. At (b) the 
inverse motive groups are accented on account of the continuous 

The style at (c) would indicate that the accented groups are 
slight extensions of the motive. Compare (a) with (c). This read- 
ing (which is according to mensural formula) is more suitable to 
passage -work than to the situation here presented. Example (d) 
is more modern and, therefore, less appropriate here. 

The C. S. is the same (in augmentation) as the first four notes 
of the motive. In order to represent these melodic groups it is 
necessary to consider them as sequence figures : 

Ex. 224. 

An additional advantage attaches to this, since it serves to dis- 
tinguish the C. S. from the subject, and enables each part to retain 
its individuality. 

After the first cadence (measure 7), the base becomes proposta, 
the partial imitations being in the treble part. The original group- 
ing is applied here : 


Ex. 225. 


Ex. 226. tEI=S: 

Measures 1 5 to 1 8 constitute a duet in canon. The two voice- 
parts are to be treated equally while they carry on their affirma- 
tive dialog. The holding tones must be accented and firmly 
sustained, and it is usually advisable in such instances to play the 
other part lightly. Observe the motive figures at the close, and 
that even this fragment 
is to be duly impressed 
upon the listener : 

The form of this invention is to be noted. There are three 
periods : the first contains seven measures ; the second (counting 
measure 7 as ending and beginning) has nine ; the third period 
contains eight. The apparent discrepancy between these curtailed, 
extended, and regular periods is less noticeable in thematic music, 
where the phrases are not so plainly outlined as in lyric works. 

The periods ending in G and in A-minor are to be indicated by 
means of a very slight rallentando and a perceptible separating of 
the closing tone from what follows. The fact must, however, be 
borne in mind that this invention is continuous, and that the 
cadences are points of relief rather than of repose. The final close 
is to be played more broadly. 

Invention IV is selected to illustrate a theory of the author which 
is frequently useful in expounding thematic music. The applica- 
tion is to continuous designs, where there is an uninterrupted series 
of rapid notes. If it is desirable (as frequently it is) to perform an 
entire passage legato, without a break in the chain of sounds, the 
necessary accents or punctuations may be left to the accompanying- 


part, usually the counter-subject. A quotation is here made from 
invention IV, beginning with measure 5 : 

Ex. 227. 

P.. I r- -p TF rT~~h g rp * F ~q 

E : lE^ g ^ I "-E^^^^E^EEE^^E3 


Since the subject consists of two measures and comprises a 
phrase, the same divisions are made in our quotation. And by 
means of primary- and secondary accents applied to the C. S. the 
phrase divisions are sufficiently indicated without interrupting the 
continuous series of sounds above. The same theory may be ap- 
plied from measures 11 to 16, where the groups of sixteenth notes 
appear below. To all such thematic passages as the one in review, 
the toccata in A by Paradisi and the scherzo in Beethoven's op. 14, 
II, this theory may be applied, and it affords an interesting relief 
from the usual methods of phrasing. 

V is constructed somewhat upon fugal lines and requires close 
analysis. The motive is this : 

Ex. 228. 



It takes precedence over every other motive group wherever it may 
occur in the thematic development. 

The same general principles are applicable to modern thematic 
works ; for example, the Allemande from D' Albert's Suite, op. i. 
Two features of the subject are to be particularly noted : 

Ex. 229. 

The broken chord (a) and the intermediate scale figure (b), together 
with their corresponding contrary inversions, 

Ex. 229. 


furnish all the material used in this interesting allemande. The 
upper part may be performed without break from the fourth to the 
last of measure 9. The last three sixteenth notes here belong to 
the following group. The lower slurs are, therefore, misleading ; 
unless an upper slur 
is used : 

Ex. 230. 

With regard to the sequences, beginning with scale figures in 
thirty-second notes, they require no particular disconnection. The 
mensural accents are supplied by the accompaniment. This is 
important here, for if the bars are thus clearly indicated, we will 
naturally follow the design until it is completed. But if the meas- 
ure becomes involved in uncertainty, we may be sure that chaos 
will result. 



With regard to the application of accent and punctuation, it may 
be stated that uniformity of phrasing is dependent upon uniformity 
of construction. When the model or the style changes, the man- 
ner of performance also must vary. J. B. Cramer's first etude is 
taken as an example in applying this theory. At first the figures 
assume this form : 

The continuance of this sequence constitutes what we may call a 
rhythmic division, and this (first four measures) should be played 
with sufficient accent to define the three elements which relate to 
time : measure, rhythm, movement. At 5 the figures are reversed : 

Ex. 232. 

Here a distinct accent is required, especially since the sequence 
descends from this point. The design is slightly varied in measures 
7 and 8, though it continues to descend : 


Ex. 233. 


While the syncopations above require some emphasis, the theme 
below must not be neglected : 


Ex. 234. 



After the incomplete cadence the sequence figure is again al- 
tered : 

Ex. 235. U3C 

This embraces three measures. The arpeggio chord, having 
been played^ already sup- 
plies the marked accent 
for this division. Another 
three-measure rhythm be- g x 235 
gins at 13, and this also 
demands one principal ac- 
cent, thus : 

Another division begins with the design in suspension : 

Ex. 237. =2= 

The last four measures are so constructed harmonically that they 
bring the etude to a natural close, especially if a well-graded 
diminuendo accompanies the cadence. Altogether there are eight 
sequence figures, and each of these constitutes a rhythmic divi- 
sion. The first series of groups ascend during four measures. Then 
there are four descending measures ; but these are divided by means 
of a change in the pattern group (see b and c), and the subdivision 
should be slightly marked. 

These rising and falling rhythms succeed one another alter- 
nately, thus simplifying the performer's task to a considerable 
extent. The student would do well to mark these outlines with a 
primary accent /\, for they should appeal on paper to the visual 
sense as plainly as to the auricular sense when the etude is per- 



The entire study is consistent and musical, though the ordinary 
phrase divisions do not appear. In other words, it is motivized. 

Melodic designs and sequence figures appear in almost endless 
variety, and frequently they counteract, by their peculiar rhythmic 
situation, the prevailing form of measure. Usually these designs 
are easily apprehended on account of the symmetrical contour. 
In such instances this symmetry must be preserved, even though it 
directly contradicts the formula of mensural accentuation. Thus 
the following from Beethoven's op. 14, II, which is copied literally 
from Biilow's edition : 

Ex.238, (a). 



Observe that the mensural accent is entirely absent in measures 
3 and 5. The melodic design or sequence figure (b) 

is really in measure, especially since the harmonic substance 
naturally falls into dual groups, thus : 

Ex. 239. 

* \ ^ * m r* j^-0- 



The modulations, from G, are to A-minor, B-minor, and C-major, by 
means of the dominant seventh chords to those keys. The accent 
marks are, therefore, perfectly logical. The effect would be similar 
had the passage been written in this manner, according to the 
method of certain modern composers : 

Ex. 240. 

There is a thematic figuration in von Weber's Concertstiick 
which illustrates another phase of this subject. The passage occurs 
at the end of the march, and consists of a series of short climaxes 
ascending higher and higher. Each of these rhythmic divisions 
contains three measures, and marks a change in harmony. The 
ground plan may be thus represented : 

Ex. 241. 

In addition to this the right hand executes a series of ascending 
figures beginning piano and ending forte. The last tone at the end 
of each passing scale figure is accented because it marks a change 
in the harmony as well as a melodic outline above. These distant 

Ex. 242. 

represent melodic climaxes, and they must be sufficiently impressed 
upon the listener to linger in the mind for a period of three meas- 


ures. The special accents have, therefore, a threefold significance : 
they indicate the beginnings of melodic and of harmonic designs, 
and mark the rhythmic divisions three-measure phrases. After 
nine measures of this thematic passage the uneven phrases are 
followed by equal rhythms of two measures each. Then there is a 
continued thesis, to be performed accordingly, without regard to 
regular rhythmic grouping. The passage quoted comprises an 
intermezzo, which connects the march in C with the presto assai 
in F t and represents the anticipated joy of the warrior-knight's re- 
turn to his lady-love. This is the meaning of the con molto agita- 
zione, the crescendo assai, the pedal-note, and the impetuous sweep 
of the joyous presto in F. 

Selections may be made from the old clavichord and harpsichord 
works. (The music of this epoch is mostly thematic.) Also, 
" Momento-Capriccioso," N. von Westerhout. 


In comparison with thematic music the lyric style stands in an 
opposite relationship. This is particularly true in regard to piano 
music. The lyric. style represents, primarily, the spirit of song. 
In order to produce a singing effect it is necessary for the pianist 
to employ a peculiar quality of pressure touch. Not the mensural 
emphases which accompany a march, waltz, or polonaise, nor the 
rhythmic accents required in thematic music. But every tone in a 
lyric melody (such as the " Slumber Song," by Schumann) is to be 
made distinctly manifest without accent, each melodic note being 
sounded somewhat equally, regardless of the particular part of the 
measure in which it may occur. That is to say, the performer 
must learn to avoid mensural accent in giving prominence to a 
lyric theme. This distinction is a prerequisite quality in the per- 
formance of nearly all slow movements, nocturnes, and songs 
without words. 

Therefore it must be understood here and hereafter that this 
pressure touch carries with it no percussive nor pulsatalic effect, 
and that stress or emphasis in the performance of cantabile pas- 
sages is primarily a condition enforced by the nature of the instru- 
ment. For example, if the principal melody in Schumann's exqui- 
sitely tender slumber song were sung by a soprano voice, there 
would be no necessity for accent. In fact, rhythmic accentuation 
would pervert the intended expression and offend the esthetic sense. 
Therefore, though considerable pressure touch may be required 
in producing a singing tone from the piano, the mechanical effect 
which is naturally associated with accent must be scrupulously 

For a number of years past the author has had an impression 
(now grown into a conviction) that the phrases and sections in cer- 



tain well-defined lyric themes are so naturally distributed and so 
plainly outlined that no punctuation or special rhythmic accentua- 
tion is required to indicate them. 

Such song-like themes as are here contemplated may be com- 
pared to short and simple lyric poems ; for instance, the " Hedge 
Rose," by Goethe : 

" Once a boy a rose espied 
Blooming in the wildwood ; 
Blushing on the thicket side 
"* He its dainty bud descried 

With the glee of childhood." 

The simplicity of expression is so natural as to leave no room 
for doubt concerning the poet's meaning, even though the lines 
were recited in the lisping tones of childhood. These simple, 
rhyming cadences require no elocutionary effort to make them ap- 
preciable. In truth, the good effect depends upon artlessness, 
rather than upon artistry or 'formula. 

Certain vocalists, forgetful of the fact that naivete is the greatest 
charm of this Goethe-Schubert ballad, proceed to embellish it with 
tempo rubato and operatic nuance. 

A fragment of lyric melody is here quoted from Beethoven to 
illustrate the present application of pressure touch : 

Op. 7. (From the largo. ) 

Ex. 243. 

The accent which, according to formula, falls upon C in the base 
at X must here be omitted, because it would detract from the sus- 
tained melody tone in the tenor part. And the note which comes 
after the second beat (usually unaccented) is here played promi- 
nently, because it is part of the theme. A similar instance is taken 
from the same composer's op. 13 : 

1 86 


Ex. 244. 


... , .-/-i-^-i^T Si*3=fct 

As the metrical accents fall upon the two equal divisions of a 
measure, we would ordinarily mark the middle of the measure at 
-f- ; and this formula would exclude an emphasis from the 
melody note, d-flat. But a higher law than mensural accentua- 
tion here intervenes and reverses the process last described. In all 
such instances no accent is to be included in the accompaniment at 
-{-. The same is true of measure 4 in this adagio.* See also 
the adagio in Beethoven's Op. 10, I, third measure ; and the D-flat 
Nocturne, by Karganoff. 

In order to maintain the song-like character of such themes as 
those under notice, the pianist should have in mind the dynamic 
quality of an organ tone. This may be represented thus : 

A lyric number will now be considered in its entirety. I. J. Pad- 
erewski's Melodic, op. 16, II, is selected. The two measures of 
prelude may be lightly accented according to our mensural formula.f 
Then the theme commences. 

A sonorous and somewhat resonant effect is intended, to which 
the motivized accompaniment contributes. Every melody note 
should vibrate distinctly (like a voice) and with more or less 
equality, the degree of force employed being influenced by the 
natural expressiveness of the theme, rather than by mensural con- 
siderations. For instance, the second quarter note in the first 
measure of the melody receives quite as much pressure as does the 

* This principle is intended to apply to the accompaniment only. A subtheme appears 
in the base, and this must not be ignored. 

f In every edition which the author has seen the measure is incorrectly marked f . 


I8 7 

note which falls upon the main accent. The same may be said of 
measures 7, 8, and 10. 

A more satisfactory effect will be produced if the first phrase is 
considered as embracing four measures. 

This corresponds to the slur drawn by the composer over the 
four measures, which are to be closely connected. The separate 
slurs in the second phrase indicate the sequence, rather than dis- 
connection. This manner of phrasing, too frequently encountered 
in " special editions," is more applicable to violin than to piano 
music. To remove all ground for doubt, the author would unhes- 
itatingly mark the phrase in this manner : 

Ex. 245. 

un poco dint in. 

The last measure is a question, and the harmony of the dominant 
is to be connected with the following tonic. 

In measures 15 and 16 (counting from the first) the interval of 
an eighth is substituted for that of the fifth, the former being 
stronger in its expression than the latter. 

The sequence is marked piano, but it is seldom advisable in can- 
tabile themes to suddenly reduce the tone from /to/. This, per- 
haps, would be better : 

Ex. 246. - 


The complete cadence carries with it a feeling of repose, and the 
poco rit. continues, therefore, to the close of the period. After a 
brief pause, the tones of this first completed period are to be slightly 
separated from the second period which follows. 

1 88 


The construction of this melodie is so regular and symmetrical 
that no particular effort on the part of the performer is required to 
indicate this fact. (One exception is noted in the extended period 
hereinafter described.) 

It may be stated with regard to the accompaniment that, though 
it assumes the importance of an obligate, its construction is such that 
a generally subdued manner of performance will not detract from 
the added charm which these harmonic parts contribute to the ex- 
pressive theme above. 

Whenever a distinctly melodic design appears in the base, it must, 
however, be brought into prominence. A quotation is made from 
measures 7 and 8. The arrangement at (a) is according to the 
ordinary notation. Ex. (b) shows the manner of performance : 

i * J I 



Such counter-themes are always deserving of notice, and the 
second example will serve to show the manner of treatment in 
similar instances. 

The second period is slightly more stimulating. Each of the 
first two phrases should be connected without a break in the 
sounds. Measures 27 to 30 are plainly divided into semi-phrases, 
and brief punctuations may therefore be included. 

It must be observed that this second period is extended by means 
of two additional measures, but these are not in codetta form. 
They are merged into the closing phrase of six measures, and 
played connectedly, as well as continuously, until the principal 
strain recurs after the quasi recitativo. The regular movement is 
resumed at 37. 

Before quitting the second period, attention is directed to the 
canonic imitation. This illustrates the simultaneous application of 
two opposite dynamic effects, and may be thus expressed by means 
of symbols : 


1 89 

Ex. 248. 

This principle can be applied to the ten following measures. 
The last period before the recollection is extended and united to 
what follows. The last five measures are ad libitum. 

A few illustrations are now taken from the largo in Beethoven's 
op. 7. Nearly all the 
isolated eighth notes of Ex - 24 9. 
the melody are indicated 
in this manner : "1" "if" 

Disconnection is here indicated in three different ways: (i) By 
the brief value of the second note ; (2) by the terminating slur; (3) 
by the staccato marks placed under the treble and base parts. 
Perhaps it may be said that the slur followed by a dot indicates a 
gentle, not an abrupt, disconnection of the tones ; but the reverse 
of this is often true. 

An experienced artist will not perform the staccato chord so 
shortly as to sound abrupt or trivial. But if the student undertakes 
to carry the design into effect according to the printed notes and 
symbols, he will inevitably pervert the intended expression. 

The tones of the second chord are to be somewhat stifled, but 
without making the disconnection so short as to sound abrupt 
or theatrical. On 1 o 

the other hand, the 
moment of silence 
imposed by the rests 
is an important ele- -g 250 
ment in the expres- 
sion. The author 
suggests the follow- 
ing method : 

1 9 o 


Sound the first chord lightly, but distinctly, with the e a little 
prominent. Then raise the dampers, and on the second beat sound 
the seventh chord with slightly augmented force. (The dampers 
must, of course, be lowered here.) Raise both hands from the 
keys exactly as the imaginary second half-beat occurs. The wave 
lines above are intended to represent the beats and half-beats. The 
object should be to stifle the sounds of the second chord, somewhat 
like an utterance choked with emotion. 

Measures 2 and 3 are similar. The remainder of the period may 
be performed according to the following, which is an amendment 
to the Stuttgart phrasing : 

Ex. 251. 

+' + * * ~ ~ * * -J k 44 * * 


A completed period terminates at 8, and surely some sense 
of repose is necessary here, for the following intermezzo has no 
melodic connection with the principal theme. Yet this last tonic 
chord at 8 is marked by the revisers staccato, as though every 
mood and feeling should be measured or cataloged, like articles 
of merchandise. 

Similar misdirections occur in almost every printed score, some 
of these being chargeable to engravers and compositors. But too 
frequently the blame rests upon revisers. A well-known instance 
maybe cited from Beethoven's op. 13. Every edition which the 
author has seen gives this interpretation to the third measure of the 


adagio : Ex. 252. 

This is so contrary to the spirit of this beautiful cantilena, and 
so unnecessary withal, that one does not feel disposed to accuse a 
musician of so gross an incongruity. But the theme has been thus 


characterized in so many editions that certain piano teachers have 
grown accustomed to this disjointed method of " phrasing" (as 
they misterm it), and believe it to be quite right and proper. If 
the engraver had gone a little farther with his abbreviated slur, all 
confusion and controversy about a simple matter of esthetics would 
have been spared. 

Another similar, instance may be found at the close of the inter- 

mezzo : Ex. 253. 

From appearances one would suppose that the E-flat is to be 
isolated. Yet according to the nature of the melody it should all 
be joined in closest connection, and the appealing character of this 
minor seventh particularly demands that the E-flat shall be blended 
with the /"below. 

Similar errors occur throughout the movement. The second 
theme, for example, is thus characterized : 

Ex. 254. 


Fortunately, there are few at the present time so credulous as to 
perform it in this hysterical manner. 

One of the principal charms of these serious lyric melodies is 
their continuous, sustained, and song-like character. But in order 
to exploit a system of " phrasing," certain annotators and revisers 
destroy all sense of continuity by the fragmentary manner in which 
they place the slurs. They seem not to realize that a phrase, or a 
rhythmic group, can be indicated (when that is really necessary) in 
various ways, without sacrificing the unity and song-character of 
lyric themes. Chopin, ever mindful of esthetic requirements, left 
us some remarkable evidences of the care he bestowed upon these 
matters. Witness the F-ininor Nocturne, No. 15. The entire first 
period is catenated by means of continuous and connecting slurs. 


The uninterrupted flow of sounds in the melody part is not broken 
even at the end of this period, which is joined to the repetition. 
Not until the second cadence (measure 1 5) is there any sign or 
symbol indicating a punctuation. The period construction is per- 
fectly regular, and the expression is to be manifested through 
nuance and dynamic effect. The period does not require punctuat- 
ing points. If it did, Chopin certainly would not have forbidden 
them, as he did by the connecting slurs. (Wagner frequently 
included entire periods under a single slur. See the Vorspiel to 
Act III, "Die Meistersinger.") 

The main theme of the adagio in Beethoven's op. 13, the 
" Melodic," by Paderewski, and Sternberg's " Night Song," op. 56, 
VI, should be interpreted in the same manner as regards the absence 
of punctuation and consequent disconnection. Hundreds of similar 
examples might be quoted in proof of this method for preserving 
the unity and continuity of regular periods in lyric style. 

There are various details and circumstances which influence the 
performance of lyric music, and most of these are illustrated under 
their several titles, such as unrelated tones, resolution, antiphonal 
groups, transition, etc. 

Furthermore, lyric melodies admit more frequent deviations from 
a fixed movement than do thematic passages, especially when the 
former are slow and in the cantabile style. (Compare the D-minor 
allemande by d' Albert with the Nocturne in D-flat by Karganoff.) 
This fact need not be dwelt upon here. 

The following lyrics are included in addition to those mentioned 
in the text of this chapter : " Melodic " in G, by Rheinberger ; 
" Fleur d'ete," by Lillibridge ; " Adieu," from Book I of Karganoff's 
"Lyric Album"; "Berceuse," by Iljinski ; "Romance," Mac- 
Dowell, op. 39 ; " Siegmund's Love Song " (Wagner), transcribed 
by Bendel or Brassin. (D.) 



In this system Harmonic Style refers to those portions of a work 
which are in full harmony, where the different voice-parts are very 
nearly the same in note-value. Plain choral music affords the best 
illustration, because the melody must be simple and form an integral 
part of each chord. The result is a purely harmonic effect, melody 
and rhythm being more or less subordinated to the sequence of 
chords and their particular relationship or tonal quality. The 
choral refrain in Chopin's G -minor Nocturne is here cited : 

Ex. 255. 

* ^ 



The simplicity of the choral melody, the fundamental character 
of the harmony, and the fact that a chord accompanies each melodic 
note all this is characteristic of what here is called the Harmonic 
Style. The theme being exclusively in the upper part (where it 
naturally predominates), requires no special lyric treatment. If it 
did, the composer would most probably have written it in this 


Ex. 256. 

lu I a* or ML' Jjj- 

__^ _^_^^ ^ _ 

=g. =^^^=^=^=^^E^ 



From this we conclude that the chords are to be sounded sim- 
ultaneously and with an equal amount of tone as regards the sep- 
i3 193 



arate voice-parts, each of which contributes its share toward the 
general (harmonic) effect. 

The entire choral refrain is played legato,* with noticeable punc- 
tuations at the end of each section and period. Afterward the 
composer has indicated still broader punctuations by means of 
pauses. This is in the style of a recollection (ad lib.) and has 
special significance in view of what follows. 

The next quotation is from E. Jambor's op. 23, IX. The 
intrada is exclusively in harmonic style : 

Slow. ^awn of Day. ' ' 

A * fl*- 

Ex. 257. 

All the voice -parts here are dynamically equal, and the notes of 
each chord are to be sounded simultaneously. The author has in- 
dicated the pedal effects and the style of performance in accordance 
with the sentiment of the music. The composer made himself 
little trouble in these matters, though the cleverness of his design 
is undeniable. The first chord is favorably arranged and located 
for the introducing of sympathetic vibrations. A gentle demi- 
staccato, with the dampers raised, will produce this acoustical effect. 
In the second measure a dissonant vibration will probably result. 
If this is distinctly perceptible, the dampers must be lowered as 
soon as the dissonating element begins to manifest itself. (This 
will depend upon the condition and quality of the instrument used.) 

The first larghetto theme in von Weber's Concertstuck is in the 
harmonic style when delivered by the orchestra. After the solo 
part enters, the harmony becomes secondary : 

* Of course, the damper pedal is to be managed discreetly, so as to connect, without 
blurring, the tones. The constantly changing harmonies preclude the possibility of sym- 
pathetic vibration effects. 



Ex. 258. 



The style here is similar to that of a vocal solo wi-th pizzicato 
accompaniment, the melody being most important. The only diffi- 
culty (and this is a technical one) consists in touching the adventi-' 
tious parts below more lightly than the theme above. 

The "National Song," by Grieg, op. 12, VIII, is almost entirely 
in the harmonic style. 1 The same remarks will apply to this as to 
the choral refrain in E-flat from Chopin. 

The middle part in Wilson G. Smith's " Vesper Chimes " pre- 
sents a contrasting example. The design is purely harmonic, and 
while the explanations accompanying the quotation from Chopin 
apply in a general sense to the example under notice, the latter has 
one peculiarity which enters into its interpretation. A fragment of 
this choral part is presented : 

Andante religiose 

Ex. 259. 


i:fr=gEr^=|JEzjj |=g== 

1 1 j | I ) & i a. 1 

i r?j<3- ' - \ , 

v 1 -^, ~^y 


The tones are to be well sustained, but not too distinct in outline 

* Press the damper pedal upon 2 and 4 ; release it on I and 3. 



or accentuation. The parallel lines below are intended to represent 
the dynamic quality of an organ tone. 

Of course, the pianist can only approximate this effect ; therefore 
the symbol is merely suggestive. 

Two more peculiarities of the harmonic style remain to be con- 
sidered : where a motive is concealed in some middle part of a 
harmonic design, or where the accompanying chords are not suffi- 
ciently attractive to merit equal prominence with the melody. 

An instance of the latter kind occurs in the allegretto of Bee- 
thoven's op. 10, II: 

Ex. 260. < 

The melody, though mostly monotonic, is predominant, because 
the harmony does not change sufficiently to claim attention. The 
principal interest centers in the resolution at the end of each phrase. 
At first the melody note is doubled in the octave below, and this 
obviates the necessity of giving prominence to that part. But in 
the dissonant combination there are four harmonic parts against the 
melody. Therefore, the upper / is to be played more prominently 
than the chord below. 

A little further on 
the motive appears in 
the middle parts, and 
this (especially in Ex. 261. 
right-hand part here) ( a ) 
is more difficult to 
accentuate separately : 


The design is evidently a continuation of the original scheme : 

The (S^/to at (a) proves to be a fragment of the motive in aug- 
mentation ; but this is duplicated below, and does not require special 
treatment, as does the a-flat resolving to g-flat in Ex. 261 (a). 

If this melodic suspension were noc duplicated in the upper left 
hand part, the design would be much more difficult of execution, 
and it is this fact to which particular attention is directed. In the 
next example the right hand alone carries a melody beneath a chord 

accompaniment : 

Karganoff. Op. 20 

Ex. 262. 

The theme being legato while the accompaniment is demi-staccato 
facilitates the execution of this design. Where the chords are sus- 
tained, and especially where the melody is in an inner part, the 
obstacles are greater, and the task of the pianist becomes magnified 
in proportion to his appreciation of the musical text. 

For similar examples see Canzonetta, by P. C. Lutkin, and 
The Flatterer, by C. Chaminade. 


Discord. The most euphonious harmonic discord is the domi- 
nant seventh chord. The first two intervals are consonant, and if 
the chord be considered componently, the third interval (five to seven) 
also is consonant. But the root and seventh form a discord, and 
this requires some kind of resolution. The discord is more notice- 
able when it is inverted and becomes a second. 

Any of the following may be considered resolutions : 

In each instance a consonant interval follows the discord of a 

It may be stated as a general principle that the act of resolution 
implies connection, and therefore the dissolving of a discord into a 
concord is to be legato. Even the resolving of an imperfect fifth, 
or its inversion, the augmented fourth, usually requires a strict 
legato style : 

Air de Danse. Gliick. 

Ex. 264. 

f * L f^~> 




I 99 

The discord should seem to melt into the consonant interval. 

The principal diminished seventh chord is classed here. Though 
the interval from root to seventh is a secondary consonance, the 
three minor thirds give it a transitional effect and place it outside 
the realm of concords. This is here illustrated : 

(a). (b). 

Ex. 265. PZ^ 

The complete chord appears at (a) with its practical root below. 
At (b) the diminished seventh and its enharmonic equivalents may 
be seen. To the ear all these are consonant, (c) shows the two 
imperfect fifths, which must be classified as discords, because the 
interval of an imperfect fifth can not represent repose or finality. 
The natural resolutions of these imperfect fifths (and consequently 
of the entire chord) are exhibited at (d) and at (e). 

The various species of augmented sixth chords are included 
here ; likewise the milder forms of secondary seventh chords, such 
as those founded upon six and seven 
of the major scale. A secondary 
seventh chord of this species (Ex. ( a )- 

266, a) has no resolution, theoretically ; 
but practically, the element of discord must sooner or later give 
place to an element of concord, thus : 


The seventh becomes a sixth and the second becomes a third, 
either above or below. 

It is evident, therefore, that the student who seeks to interpret 
music -must understand the general principles of resolution. 

Dissonance. Briefly stated, this consists of a very harsh inter- 



val (as a minor second or a major seventh) or a combination of two 
discords. And it may also be stated, according to the author's 
theory, that the act of resolution and the degree of connection 
increase in importance -with the degree of dissonance. Following 
are illustrations : 

The dissonant ninth resolves in strictest legato to the octave, and 
the seventh below is connected with its natural resolution, the sixth. 
Here the tones are disconnected. Therefore the f-flat above and 
the d-flat below require particular connection. 

The next quotation is similar, except in regard to movement : 

Ex. 268. 


Grieg. Op. 12, III. 

The d-sharp at (a) forms a discord with c-sharp, and a dissonance 
with e. A partial resolution takes place when the d-sharp descends 
to b. The final resolution occurs at (c), and here the phrase is 
separated from what follows. At (d) the discord and the disson- 
ance both disappear when d-sharp resolves to c-sharp. The 


2O I 

dissonances require the closest connection, though all the resolutions 
are indicated by means of short slurs. A series of discords and 
dissonances are here presented. These illustrate more plainly the 
foregoing principles : 

Ex. 269. < 

This is essentially organ music, and all the connecting notes are 
supposed to be tied. The slurs, therefore, show the resolutions 
and progressions which require, particular connection, legatissimo. 
The same formula applies to dissonant passing tones (a) and to 
appoggiaturas, (b) : 



Ex. 270. 

sp=titig= [==|id=it=q 

It is simpler to call the d-flat in the base an appoggiatura, rather 
than an inverted ninth. But in either instance the dissonant tones 
are to be strictly connected together. 

A double discord is to be considered here, since the two discords 
combined produce an effect similar to that of dissonance. An in- 
stance has already occurred in Ex. 269. The discord of suspen- 
sion at -{- contains two intervals which require separate resolutions. 
It is the same here : 




Ex. 271. - 

Owing to the suspension of d, there are two discords in the com- 
bination at (a). These, are resolved separately, as may be seen. 
The following double appoggiatura is resolved simultaneously : 


Ex. 272. 

=g=g qi:=i=j 

The discord e-flatf, and the dissonance b-flat a, are both 
connected in their resolutions by means of strict legato, though 
under ordinary conditions the music of this number does not 
require smoothness of execution. Thousands of examples similar 
to the following from Chopin might be cited : 


Op. 37, I. 

Ex. 273. < 



The discord at (a) resolving down, and the dissonance at (b) re- 
solving up, are played legatissimo. 

A few instances will now be presented in which the discord is 
disconnected from the following concord. This naturally presup- 
poses that the prevailing style is staccato. And it may be added 
that such examples usually contain the milder forms of seventh 
chords, thus : 

Ex. 274. 


Th. Kirchner. Op. 7. 





This is so simple as to require no further comment. 

A more unusual example is the ending to the allegro in Bee- 
thoven's first F-minor sonata : 

Ex. 275. 




Only three of these chords are consonant, and one is dissonant. 
The passage occurs in the form of a stretto, and represents a very 
impatient, not to say petulant, mood. The nature of the sounds 
and the allegro movement combine to make this less exceptional 
than it would seem to be. Also, the progression of the last two 
chords in the base is such that no resolution could possibly take 



place, and this circumstance is to be borne in mind whenever a 
similar instance presents itself. 

Long appoggiaturas almost invariably come within the rule of 
discord connected with concord. Since the harmonic appoggiatura 
is foreign to the accompanying chord, we may reasonably conclude 
that the former requires a legato style, as do nearly all resolving 
discords : 

Yon Weber. Op. 79. 

Ex. 276. 

The melody notes which fall upon the principal accent are har- 
monic appoggiaturas ; and though the prevailing style is staccato 
(in keeping with the joyous mood), the appoggiaturas are dissonant, 
and, therefore, closely connected in their primary resolution. 

Such instances are so numerous that it seems unnecessary to 
quote further. 


It is desirable to consider first the general impression or total 
effect of certain harmonic progressions. Unless the harmony 
serves as a mere framework to a lyric melody we are inclined to 
associate with every chord progression some definite melodic out- 
line. Every changing harmony necessarily involves a melodic 
progression in some of the harmonic voice-parts ; and since our 
tuneful instinct is naturally developed, we unconsciously follow the 
single thematic progressions, rather than the effect of relationship 
between one chord and another. 

Harmonic movements may be considered in their relation to 
sequence, cadence, transition, or polyphonic coloration. Harmonic 
sequence embraces some particular form or arrangement of chords 
which is repeated in the free or the strict style. The original 
design, or pattern group, assumes the importance of a regular 
motive, and in the sequence every repeated figure will require some 
form of accent or punctuation. The harmonic sequence is, there- 
fore, a law unto itself. The following excerpt shows one phase of 
the subject : 

Arthur Foote. Op. 30, I. 


H^HH-^ ^ ' 

Ex. 277. 

This model is indicated by the first two chords : a triad in its 
first inversion, followed by a triad uninverted. Observe that each 
repeated figure is marked by the composer as a separate group. 




The disconnected bars indicate the accents during the continuance 

of the sequence ; that is 
to say, each group should 
be marked. It is some- 
what remarkable that there 
are four melodic parts em- 
braced in this harmonic 
sequence, each melody 
being distinct and individ- 
ual : 


gizzzzznpz ' 

Ex. 278. 



Yet it is the total harmonic effect which must be represented, 
for the ear can not be expected to distinguish four themes simulta- 
neously, unless they had been heard previously in an isolated form. 

Attention is likewise called to the dominant relations in this 
sequence : E-flat A-flat, D G, C F, etc. See Ex. 277. There 
are other sequences in this opus, which the student should discover 
for himself. 

There are several elements of transition to be considered here. 
The subdominant and leading note (to any major or minor key) are 
usually most important. These are well-known, but they must be 
understood in all keys. These transition elements form the third 
and seventh of a dominant seventh chord, and the root and fifth of 
a diminished seventh chord, thus : 

Ex. 279. 

The two principal resolutions of an essential seventh discord are 
shown at (a) and at (b). The subdominant and leading note are 
also contained in. the diminished chord at (c), but these elements 
are now represented by a-sharp and e. In every instance the 


resolution is to the tonic and third of the resulting concord. The 
minor sixth (g in Ex. 279, c) also is an element of transition, 
and therefore connected in its resolution. To these must be added 
the dominant ; and in diatonic modulation the dominant relation 
operates so naturally and potently that it is possible to outline a 
series of modulations by means of fundamentals alone : 


These are sufficiently 
inciting to suggest the 
following harmonies : ( b). 





The student may complete the example according to this sequence 
model, and end in C. 

The augmented sixth chords, i, 2, 3, are transitional in their 
nature, No. 2 being particularly strong and bold. The interval 
which gives to these chords their specific name can not occur 
naturally in any major or minor key. Therefore is their tendency 
transitional : 

Ex. 281. 

===i==Fq= ^ 

All these chromatic tones are to be emphasized. Augmented 



sixth chords are rather strong and bold, and this expression fre- 
quently counteracts the effect of smoothness which so many upper 
and lower leading tones would seem to indicate. Thus, from 
Beethoven : 

Ex. 282. < 

This is a No. I which almost invariably resolves to the second 
inversion of a minor concord. 

The secondary seventh chords noted in Chapter XXII are non- 
transitional, and though they frequently require a strict legato style, 
they do not, on their own account, demand accentuation, as transi- 
tion chords usually do. The secondary discords are used princi- 
pally as connecting links between the consonant triads and the 
principal discords. 

Also, the preceding must influence us in determining upon the 
points where modulations and cadences are effected. With this 
understanding we undertake a brief harmonic analysis of the well- 
known C-major prelude from Bach's " Wohltemperirten Klavier." 
In this, and in all similar works, the performer should be influenced 
chiefly by harmonic considerations. The prelude is in broken 
chord form, sixteen notes to the measure. Only the harmonic out- 
line is here presented : 

Ex. 283. 


This complete cadence-group is easily understood, and it is so 



natural as to require no special treatment. The second phrase also 
has four measures ; but the third phrase has only three. These 
seven measures should be catenated, so that the unequal section 
will not be particularly noticeable : 


Measures 7 and 8 form a sequence of 5 and 6. These are reg- 
ular and are accented accordingly. The suspension at 8 serves to 
link the harmonies together, and this will assist in preparing the 
cadence on G. Observe the rallentando and accent at 10. 

A series of natural modulations by means of diminished seventh 
chords then follows : 

Ex. 285. 



The transition tones are to be distinctly marked, and this will 
serve to outline the rhythmic form. The resulting concords are 
inverted, which prevents the effect of a terminal cadence. 

Another prepared cadence follows, and this is easily managed 

because it is natural and regular : 



Ex. 286. 




The remainder is a continued thesis of 17 measures, and this 
requires considerable dynamic effect. The impression of a com- 
plete cadence is not to be created at 26, because the dominant 
pedal-note continues beyond this point. The maximum amount 
of tone is developed at 30, and from here a gradual diminuendo is 
to be made, accompanied during the final cadence by a slight ral- 
lentando. The continued thesis at the close will, if played unin- 
terruptedly, sufficiently justify the irregular period construction. 

In strict designation, the last four measures comprise the coda. 

Under this heading is included the influence of a fundamental 
base in terminal cadences. The following will serve to illustrate 

this point : 

Hans Huber. 

Ex. 287. 


M Li J 

If the fundamental be discontinued after the first beat, the final 
chord will seem to be out of balance ; there will be nothing for the 
harmony to rest upon at a point where the ear demands repose. The 
composer undoubtedly intended the fundamental to continue in vibra- 
tion, but he does not say so. The effect should be like this : 

Ex. 288. 

L. H. 



^-H &>-: ~l:2_irJ 


21 I 

The sostenuto and damper pedals should be employed here, 
though a tolerably satisfying effect can be produced with the 
damper pedal alone. 

Numerous instances of this kind occur in piano music, some of 
which are less satisfactory than that quoted from Huber. If, how- 
ever, the ear be consulted, a remedy will readily suggest itself, as 
in Ex. 288. 

Somewhat similar is the following from Rubinstein's Taran- 
tella : 

Ex. 289. 

This introduction is founded upon a dominant pedal-note, and the 
effect of this should prevail throughout. Otherwise the chord at 
-f- sounds incongruous on account of its want of connection. 
The pedal-note may be maintained by raising the dampers, or it 
may be played in this manner : 

Ex. 290. ^ 

In either case the effect will be more satisfactory than it is accord- 
ing to the original notation. 

With regard to so-called " accidentals," it is necessary to distin- 
guish between chromatic passing tones and those which directly 
affect the tonality. Knowledge of harmony is here necessary. 

An excellent preliminary example is the modulatory Prelude by 



Beethoven, opus 39, II. It begins in C and passes twice through 
all the major keys by the normal fifth relation. The aim is purely 
modulatory, though the prelude affords a very good illustration of 
logical thematic development. Nearly all the modulations are 
effected by means of the leading note to the next key in order. 
The dominant plays a more important part than does the sub- 
dominant. The first transition tone to be marked \sf-sharp in the 
contralto part, measure 5. This /-.y/zar/, together with the dominant 
below, establishes the key of G. It is so with c-sharp at 9, g-sharp at 
10, and d-sharp at 1 1. These are the leading notes to D, A, and E, 
and even where they occur below they are to be duly marked. 

Measures 12, 13, 14, and 15 illustrate what was said about chro- 
matic passing notes. The tonality is quite the same on the second 
beat of measure 16 that it was at 12 i. e., four sharps. There- 
fore the chromatic passing notes in measures 14 and 15 are to be 
considered en passant ; they were not intended to create any par- 
ticular key impression. But the a-sharp at 16 must receive special 
treatment because it leads to B at 1 7. 

After passing to C-sharp at 21 the enharmonic equivalent (D-flat) 
is substituted, and the modulations by fifths continue. Otherwise 
the cycle would have ended in B-sharp major. 

The leading note continues to be the principal element of transi- 
tion, and the "accidentals" at 35 and 36 require no special 
accents. But on the last beat of 37 the e natural passes out of 
B-flat into F y and must be marked. These leading tones are to be 
well connected in their resolutions. More chromatic passing notes 
appear between 38 and 45, but since the object is to reestablish 
the key of C at 46, the last measure claims particular attention : 







A strict legato style in the final resolution assists in perfecting 
the cadence on C. 

The remaining cycle is similar in theory though different in 
treatment, and the student is advised to examine the second part 
with a view to distinguishing between the chromatic passing tones 
and those which terminate the various tonal divisions. 

M. Lussy was the first author who attempted to formulate a 
rule in reference to transition tones. He gave directions that every 
transition tone must be accented in order to enforce upon the ear 
the new tonality, especially where the transition is remote or unex- 
pected. But a distinction is to be made between forced accentua- 
tion and distinctness of utterance. And furthermore, the nature 
of the music must exercise a governing influence. For instance, 
there is a section in Elsa's Dream (from "Lohengrin") which 
passes from A-flat through C-flat y G-flat, F-sharp minor, and 
A-major, all in the course of a few measures. The sentiment is 
such as to exclude strong accentuation, and yet the performer must 
unfold the distant transitions so that they will be distinctly appreci- 
ated by the listener. Nothing could be more appropriate to the mood 
here represented than these mysterious enharmonic transitions ; 
but this circumstance increases the difficulties of interpretation. 

Transition usually represents some kind of emotional excitement, 
determination of purpose, or changing scene. Certain modulations 
are bold and positive ; others have a progressive, onward tendency ; 
and there are modulations which are reminiscent and seem to 
recede. As these present themselves their esthetic qualities will 
be noted. 

From Grieg's op. 37, I, an instructive example is taken : 

Ex. 292. 





This mode of transition is peculiar to the great Norwegian com- 
poser. Despite the word " tranquillo " it will be necessary to 
strongly mark the e in measures 2, 3, and 4, as this is the connect- 
ing link in a somewhat rugged chain of harmonies. Also, c nat- 
ural in the fifth measure requires a decided accent, because it deter- 
mines the transition to F. This section is repeated in sequence 
form a half step higher, and then there are 14 measures of stretto. 
These chromatic transitions are very inciting and forcible on 
account of the augmented 6th cords No. 2, the strongest and 
most impelling combination known to harmonists : 

Ex. 293. 

S~~~\+ L h 

*+-*+- i *-I#*- * -* J'U^-^-^&H^^'-I ! 

^ I i j j I ^^^^^J i M I I 

A continual increase in tone and movement must accompany 
this exciting passage, and the augmented 6th chords above and 
below are to be forcibly accented. The dynamic signs and sym- 
bols are those of the composer. 

An equally important esthetic consideration is that of harmonic 
coloration, though to the performer this is often vague and intan- 
gible in its influence upon theoretic interpretation. Reference is 
here made to certain chord relations, and to chromatic intervals 
which do not materially affect the general key-impression. In the 
latter instance the object is to impart a particular color to the 
music, rather than to decide a given tonality. Such is the effect 
of Schumann's great quintet, op. 44. An air of seriousness and 
mysticism envelopes the music, because the key-impression seems 
to be continually changing. Instead of accenting the chromatic 
tones, we should rather consider them as belonging (even though 
somewhat remotely) to the main key, and thus preserve the spirit 
of mysticism which the composer infused into the music. Through- 



out the entire opus the harmonic coloring is of this vague, romantic 
character. Not until the composer has .evolved out of previous 
motives the bright carillon theme does the purpose reveal itself. 
It seems like dwelling in a mountain cave until that which appeared 
like formless outlines is finally recognized as symmetrical columns 
and delicately carved arabesks, all the more beautiful when 
screened from the light of external illumination. 

In the following example from Rubinstein it is the chord rela- 
tions which are to be noted : 



Ex. 294. 

The A-minor chord has a peculiar effect here, in keeping with the 
oriental character of the music. Hence the accent on the last 8th. 
Also, observe the harmonization here : 

Ex. 295. 

Still more suggestive are 
the various harmonic ar- 
rangements of this Asiatic 

theme, especially where Ex. 296. \~~^~^~~ 

the augmented triad is 
used later. 





In the following excerpt from von Weber the characteristic tone- 
color is imparted by means of altered intervals in the base : 

Op. 79. 


Ex. 297. 

i j^JE^^EJg|E 

This occurs in F-major, and therefore there is no rule that calls 
for d-flat and e-flat in the solo part. Yet how much more expressive 
this than the following I. ^^ -^ 

ment : 


D natural, as part of the essential discord above, resolves as freely 
to c as does the d-flat ; yet the latter gives an entirely different 
meaning to the passage, and influences the performance accord- 
ingly. The following phrase is similar : 


Ex. 299. 


The D-flat is used in place of d natural not for a transitional pur- 
pose, but because the former is a stronger, darker tone. 


Gavotte, Karganoff, op. n. A natural in the seventh measure 
is the first transition tone to be marked. Then f natural and b 
natural at 8 must be made prominent. A-flat in the musette is 
only a passing tone. 

Nocturne, R. Tempest, op. 2, II. Minuet in , L. Godowsky. 
These selections are about grade 4. 


Ad libitum, obligato, and polyphonic accompaniments are 
included under this general title. Indeed, the variety is so great 
that the author is inclined to begin their consideration with this 
preliminary statement : The more uninteresting an accompaniment 
may be, the less notice should it attract in the performance. In 
other words, the adventitious parts should be subdued and kept in 
the background in proportion to their lack of musical interest. 
From this general statement we may proceed to consider various 
styles. A quotation from Chopin is presented : 

Op. 64, I. 


This left-hand part represents no particular design other than that 
of going with the solo, to which it serves as harmonic and rhythmic 
basis. The accompaniment is, therefore, to be played in a very 
subdued manner. Such instances are of rare occurrence in the 
works of Chopin, and even in this waltz the base part of the second 
period is musically much more important. 

The forty-seventh mazurka affords another instance in which the 
accompaniment is, in itself, uninteresting, and, therefore, unimpor- 
tant. The same may be said of the tenth mazurka. 

Isolated examples may be found among the waltzes and mazur- 
kas, and all such are to be treated in the manner described. But 
usually the accompaniments of Chopin represent some musical de- 
sign, either harmonic, melodic, or rhythmic, and these must be 




treated with the greatest care and intelligence. It will here suffice 
to mention the op. 37, I and II, and the op. 72, I. 

Nearly all forms of common chord accompaniment require a 
subdued style of performance ; and where the same harmony is 
reiterated, it is frequently desirable to employ the slurred staccto 
touch : 


Ex. 301. 

The pulsations of the chord should be so gentle as to suggest 
almost a continuous tone. 

The intermezzo in a favorite adagio from Beethoven is similar in 
effect : 

Ex. 302. < 

-+*+--*--*- -0- -0- 

The middle accompaniment to the second theme (A-fiat minor) of 
this adagio should be performed in the same manner, though it is 
simply marked staccato. But a staccato effect in such instances 
is too sprightly, and attracts too much attention to a merely adven- 
titious part. This is a fault frequently committed, and even in an 
allegro movement, such as the following, the repeated chords sound 
commonplace if played in the ordinary staccato style, as marked 
in the Stuttgart edition : 


2I 9 


Op. 2, I. 

Ex. 303. <> 

This accompaniment is uninteresting, and for that reason it 
ought to be played in a more concealed manner. The staccato 
merely directs attention to an unimportant and unattractive feature 
of the music. 

Where the lower parts, together with the theme, constitute 
a design in the har- , L Grieg, 

monic style, the 

former are not to be 
understood as mere 
ace ompaniments. 
This is illustrated 
here : 

Ex. 304. 

rr^Hi 1 I H 1 

Since each part embraces a melodic motive, they all are to be 
considered equally, as voice -parts. It is so with all similar designs 
in harmonic style. See examples 255, 257, 259, 275, 277. 

Melodic outlines frequently appear in the base without any ap- 
parent indication of their existence or their relative importance. 
Such an example is quoted : 


P. Douillet. Op. 7. 




A distinction is to be made between the gradually ascending 
base (which is melodic) and the middle harmonic parts. The latter 
are subordinate, and require no special treatment. The design of 
the accompaniment may be thus represented : 


E*. 306. 








_ I 

if2_i : 

This is particularly effective here owing to the thematic character 
of the treble part. In this valse caprice there are a number of 
these hidden melodic designs (ascending and descending scale-wise), 
and it is the student's task to apprehend them and perform them 
accordingly. A similar instance occurs in the following, though 
the sub-theme below is more plainly indicated here : 


Ex. 307. 

___ F __,_- 2^-7 

L -U i P h^ * 


The counter melody in the base must be duly impressed upon 
the listener. 

Progressions like these may occur in any of the accompany- 
ing parts, and they are always to be treated melodically. Thus, 
from the same opus, two inner voices require special treat- 



Album Leaf I. 

Ex. 308. 





The syncopated motive in the two middle parts requires a gentle 
but perceptible accent, the pedal-note base being the most subdued. 

Harmonic designs frequently appear among the subordinate parts 
without any distinctive melodic outline. Following is an example : 

Ex.309, (a). 

Beethoven. Finale Op. 2, I. 

~^ A i -^ i -i ' i~3 ' ' ^ i i-^-k-i !- i ni> i~l ^-i ^ 1 

&be=pdbj- J r J ^--g3fc=^^szggiri=3 

This harmonic sequence appears somewhat in form of suspen- 
sions. The design to be made manifest by the performer may be 
thus outlined : 

No particular accentuation is required in the left-hand part, yet 
the effect of the sequence must not be wholly obscured by the 
more positive theme above. This is not a simple task, and there- 
fore is the student advised to perform the left-hand part according 
to Ex. 309 (b) until this effect can be transferred to the original 
sequence, (a). 



Sometimes the accessory parts are so complete in themselves as 
to constitute a distinctive feature of the music. Thus, from a 
Canzonetta by V. Hollaender : 

p< m 

Ex. 310. 

This style of accompaniment continues, somewhat in the mannei 
of a ground-base, from first to last. It is so musical and so char- 
acteristic in design as to demand special treatment. Such accom- 
paniments are more than the framework to a picture ; they consti- 
tute the background, and are, therefore, part of the picture. The 
left-hand part should seem to issue from some other instrument, as 
a harp, or perhaps a guitar and mandolin. The main theme above 
is, of course, to be played cantabile. 

Another style, somewhat analogous to this, is represented in the 
next quotation : 

Ex. 311. (a). 

A.ndantino con moto. 

Bavarian Song. J. Gibsone. 

The graceful arabesk figures in the middle parts are almost as 
interesting as is the theme ; and though they are to be performed 
lightly, the fact must be noted that owing to the difference in 
rhythmic arrangement between the measured notes of the melody 
and the sixteenths of the accompaniment, the latter are mostly 
heard alone, and may, therefore, be extremely soft without losing 
their identity. Inexperienced performers will find it useful to read 



the accompaniment first (with very little accent), omitting the 
theme, thus : 


The same remarks apply to the accompaniment in Mendelssohn's 
song without words known as the Duetto. Likewise to Bendel's 
Silver Spring, from op. 137, and the Nocturne in C-sharp minor 
by Karganoff, op. 18, II. 

Spinning songs, boat songs, cradle songs, require special atten- 
tion with regard to the subordinate parts, because the accompani- 
ments to these single forms are intended to be suggestive. Any 
standard barcarolle or spinning song will serve to illustrate this 
point. With regard to the former, the accompaniment is supposed 
to suggest some, regular, undulating form of motion, like that of a 
boat upon the water. This requires considerable dynamic effect, 
yet without a suggestion of angularity. An example is quoted 
from Loeschhorn'.s "A Venise" : 

^- x^^ 

f I 

Jtr9 t i5:i^ : ^E: :=)=! 

The evident intention during these preludatory measures is to 
suggest, in a relative way, the rocking motion of the gondola. 
Then the theme begins. This seeks to express (with aid of the 
accompaniment) the scene and the mood, rather than the cantilena 
of the gondoliers. The song occurs in part II. In music of this 
genre the accompaniment becomes more prominent than it ordi- 

22 4 


narily would in a romance or a nocturne. (See the barcarolle by 
Chopin, op. 60.) 

In the spinning song, also, considerable importance attaches to 
the attendant parts ; indeed, the outward, symbolic character of the 
work is reflected in these. It will be sufficient to mention these 
instances : Margaret at the Spinning-wheel, Schubert ; spinning 
song from Wagner's " Flying Dutchman," arranged by Liszt; 
La Fileuse, J. Raff; " Le Rouet d'Omphale," by Saint-Saens, op. 
3 1 ; Spinning Song, Otto Hackh, op. 50. 

A brief example is here quoted from Saint-Saens, op. 3 1 : 

Ex. 313. 

These rapid figurations typify the sound and motion of a spinning- 
wheel, and therefore the general character of the scene. The 
pianissimo symbol does not signify that the accompaniment is un- 
important, but the effect is much more suggestive when the figures 
are played lightly. Of course, it is presupposed that the dynamic 
quality of the accompaniment will increase or diminish with the 
varying expression of the theme. 

Music suggestive of the chase is usually founded upon a wald- 
horn motive ; and even though this may be confined to the accom- 
paniment, it should be treated prominently. A characteristic ex- 
ample occurs in the overture, "Genoveva," though this is a prin- 
cipal theme : 


Ex. 314. 

There are many of these hunting calls and signals, the one from 
Schumann being fairly representative. Composers use them freely, 
like quotation mottoes, or as sculptors employ the symbol of 


balanced scales in their statues of Justice. See La Chasse, by 
Heller or Rheinberger ; the Hunting Song in A by Mendelssohn, 
or R. Franz' song, In the Forest. 

Carillons also require special treatment when the bell motives 
fall among the accompanying parts. The analysis of Field's 
" Midi " rondo shows the importance of understanding the carillons 
and their various ramifications. (See Chapter XIII.) 

Those accompaniments which are seemingly unimportant fre- 
quently present the greatest difficulties to young performers. The 
nocturne in D-flat by Chopin may be cited as an instance : 

Ex. 315. Op. 27, II. 

The tender grace of the theme is so gentle as to require a zephyr- 
like touch in the left-hand part. These soft, murmuring accom- 
paniments are usually played too prominently, except by great 
artists, who understand that a mere rhythmic outline or shadow- 
like -background is intended. In a general sense, these remarks 
apply to No. I of the same opus. It is scarcely possible to sound 
these accompaniments too softly. (See also the nocturne, La Fon- 
taine, by A. Henselt, op. 6, and the Song of the Brook, by Lack.) 

One of the most important and effective obligate accompani- 
ments is that to Chopin's Etude, op. 10, XII. Its importance may 
be said almost to exceed that of the fragmentary theme above. 
The portentous conflict in the left-hand part pursues its irresisti- 
ble course with but little regard to the declamatory motives above, 
which may be compared to the voice of a general giving his com- 
mands while the battle continues to rage. 

Equally difficult, though less turbulent, is the accompaniment to 
Chopin's nocturne, op. 32, I. It contains regular designs, sub- 
themes, and canonic imitations, all of which require particular and 
separate treatment. 



Polyphonic accompaniments must now be considered. A large 
quantity of the harpsichord and old organ music is in this style, 
i. e., the accompanying parts are vocal, or at least independent in 
their melodic movement. An example is quoted from D. Scarlatti : 

Ex. 316. 


(g^g^^^ E ^^^^=^i^TT 

Here are three independent voice-parts, each requiring a different 
manner of performance. The two parts in suspension must be 
played in the strictest legato style, with a vibratory accent upon 
the part in syncopation, so that the resolutions will be plainly 
audible. The lowest part, though it is a sequence of the principal 
motive, may be played non legato, in contrast to the scale melodies 

A modern instance may be cited from E. d' Albert's Suite, op. I. 
The allemande contains scarcely any ad libitum or harmonic parts. 
The base is mostly a counter-subject, and contains fragments of the 
chief motive, either direct or inverted. There is, indeed, very little 
distinction, as regards melodic importance, between the various 
voice-parts. Also, see the Prelude by Chopin, op. 28, XXI. 

Since polyphonic accompaniment culminates in the fully de- 
veloped fugue, we may leave further discussion of this subject to 
Chapter XXVIII. 



Indications of ad libitum and a tempo. The style or form 
of an accompaniment frequently serves to indicate in a general way 
whether the movement is strict or free. Perhaps this can be more 
plainly illustrated with a song than with an instrumental work, 
though the principle applies to all kinds of music. The first quo- 
tation is from Rubinstein's song, Good-night The music of the 
first two lines is quite free in movement : 

" Good-night ! good- night ! and is it so ? 
And must I from my Rosa go ? " 

Then there is a section in which the movement is more strict : 

" O Rosa, sing good-night once more, 
And I'll repeat it o'er and o'er, 
Till the first gleam of dawning light" 

(Here the strict movement is relaxed :) 

Ex. 317. 

ad lib. 

Shall find ... us say - ing still, ... good- night." 

The only section which is to be sung a tempo is that where the 
accompaniment falls into triplets, thus : 




Ex. 318. 

O Ro-sa, sing good-night once more, And I'll re - peat it o'er and o'er, 



These triplet figures are a sign of motion (if not of agitation), 
and this can not be ignored. It is so in each of the three verses. 

A similar example is quoted from " I think of thee," by Lassen. 
While the rhythm of nine eighths is maintained in the accompani- 
ment, the movement does not materially vary. But at the close 
the rhythm of eighth notes is discontinued during two measures, 
and these are a piacere : 

Ex. 319. 

-13 U^- fV-. 

^ ^ 

"^ /-ss 

~T i 

-I 4 

* 1 

j 1 1 ^ ftr-l 



Y&- -j 


dar - 

ling, wen 

thou near! _ 

t41 - r*r~~ 

1 y i 1 f~~\ 

Jtfe^ * 

^=^ r 

--^r~ ^ P r- 

\(CYV.. i 

6 Mi r>4 * 

j ^ 1 

. i y 

At the end of the vocal period the rhythm of nine eighths is 
resumed, and therefore the regular movement also is resumed. 

The application is similar in music for instruments. Observe, for 
instance, the coda to Chopin's nocturne, op. 32, 1. This was marked 
by the composer, a piacere. 

Also the first part of the A-flat Ballade, where the solo is accom- 
panied by a few ad libitum chords. The tempo here is free, be- 
cause there is nothing to indicate a fixed movement. 


But afterward, when the rhythm of the second theme begins to be 
agitated and to excite motion, the movement becomes more lively 
and more regular. Another instance from the same composer is 
his nocturne, op. 55, I. All of the main theme, with its ad libitum 
chord accompaniment, is rather free in movement. But in the finale 
the uninterrupted triplet figures are a notice to the performer that 
the regular movement is not interrupted. The only deviation is 
an accelerando at the close. 

Music in the style of A. C. Mackenzie's " Reminiscence " is 
almost invariably to be played with varying movement, either ad 
libitum or tempo rubato. Whereas the AgitatotrQT& d' Albert's op. 5- 
admits but little deviation from the fixed rate of speed. The rapid 
figurations in the accompaniment create a regular movement, and 
are, therefore, a sign of uniform (but not absolute) motion. See 
also the first of Liszt's " Love dreams," and for a contrasting ex- 
ample, " Night has a thousand eyes," by Ferdinand Dewey. A 
very good simple illustration is the Canzonetta in A-minor by N. 
Gade, op. 19, III. 

Altering of Note-values. Tempo Rubato. It is not always 
possible to represent the precise time-value of every note, especially 
in lyric themes. Nor is it desirable always to give each note its 
actual value according as it is notated. The musical pitch is, of 
course, fairly represented ; but the rhythmic arrangement is fre- 
quently left to discretion, as in vocal recitative. Thus, in the No. 5 
of Grieg's Lyric Pieces it is not presumed that every note -value 
will be preserved exactly as written. A phrase is quoted : 




Ex. 320. 

--il 1 - 1 -i - - - i T 

The last two measures 
might be thus altered : 

This is particularly appropriate in the last measure, where the 
passing note, b, is substituted for a : 




9 m- 

But in lively music of this character the usual object in altering 
note -values is to present variety to a regular rhythmic arrangement 
of melodic notes. Nevertheless, Ex. 320 (b) may serve as a pre- 
liminary illustration of tempo rubato. 

Where dotted notes are thus interpolated they must correspond 



to the nature of the music, since this rhythm 

much more sprightly and animating than the quiet, even groups, 

I i i i . See also Grieg's op. 6, II, and op. 12, VI, second 
theme, in D-minor. 

A contrasting example is quoted from Chopin's op. 55, I : 

Ex. 321. (a). 


(b). I 

As an amendment to the fourth 
measure this arrangement is offered : 

(i) Because this is a repetition of the first phrase; (2) on ac- 
count of the small note, which is here more of a grace note than 
an acciaccatura. 

These effects are produced without altering the movement ; and 
since they occur in the melody part only, it must be understood 
that the accompaniment proceeds regularly. Instances of this 
kind frequently occur in the performance of lyric music ; but the 
liberty does not apply to thematic works. 



The following quotation from a pastoral melody will serve to 
illustrate a different phase of this subject : 

Ex. 322. 

Gregh. Op. 32. 

M f I .- - I . 


It would be a very perfunctory task for a performer with musical 
temperament to play the melodic groups (a) and (b) exactly as 
written, and yet the manner of notating these groups could 
scarcely be improved. 

Those who listened to Mr. Paderewski's interpretation of Chopin's 
G-major nocturne will not soon forget the peculiar vocal expression 
which was imparted to the second theme, especially here : 

Op. 37, II. 

Ex. 323. 

There was no perceptible altering of note-values, though the c was 
slightly prolonged and blended with the a. The effect is here 
classed as a vocal one, the slight retard being necessary to the 
portamento from c to a. 

Upper and lower leading tones, when closely connected, fre- 
quently require a tempo rubato style of performance. In such 



instances it is usually desirable to slightly prolong the value of the 
first note of the resolution, whether it ascends or descends. See 
No. IV of Grieg's Poetic Tone Pictures ; also his Arietta, op. 1 2, I. 

It is in such situations as those enumerated that the tempo rubato 
serves an artistic purpose. But it would better be omitted entirely 
than abused or caricatured, as so frequently it is. 

The Carnival, by Schumann, op. 9, is a very instructive work to 
study in connection with this subject. These sketches are purely 
impressionistic, and, while the quality of the music is excellent, the 
quaint scheme of the composer is so well represented as to present 
a raison d'etre for the various effects of nuance and rubato which are 
required in the performance. See also the Hungarian Rhapsody in 
G-minor, by Heinrich Hofmann. 

The scherzos, ballades, and nocturnes of Chopin contain many 
interesting examples of tempo rubato, but those who are capable 
of performing these poetic inspirations need no advice from the 

Notation Signs. In vocal music when several notes are joined 
together by means of bars the singer understands that one syllable 
is vocalized to as many notes as are so joined, and when hooks are 
used instead of a bar (or bars), it is a sign that the music is syllabic 
one note to each syllable. A brief example will suffice : 

Loud let your voic- es sound ! let all glad an-thems raise. 

Instrumentalists do not understand the notation signs in this 
manner. They consider the hooked note as a sign of isolation, 
and this is the usual signification in music for instruments when- 
ever hooks are substituted for bars, as here : 




The isolated note is, of course, detached. The effect would be 
similar had the composer marked the C -sharp staccato. Still 
another mode of representation would be the following : 








A little further on the composer employs a different method of 
representation, thus : 

Stcherbatcheff. Op. 8, IX. 

326 . 

The resolutions of the leading tones (f-sharp, g JK, etc.), which 
must be very legato, justify this peculiar notation. The applica- 
tion is similar in Chopin's nocturnes, V, XII, and XIII. 

A curious instance is here quoted from the music-drama, " Sieg- 
fried." The example is instrumental : 

^^^ Wagner. 

.Ex. 327. 


The hooked notes here must signify something more than isola- 
tion, because every note is detached by means of staccato. 

A degree of importance attaches to the notes g-flat, b-flat, d- 
flat, which form the outline of a chord motive. Therefore, this 
reading would seem to be correct : 


i tr~r trs 1 ir- 1 Tn ^ 



Perhaps this shows 
tage the groups : 

to better advan- 

A somewhat different application of hooked notes is to be made 
in such works as the Romanza in P by Clara Schumann, op. 21, 
II. A soft pizzicato effect is intended throughout. Where the 
bar is used, it is for a resolution which must be connected arco. 

Separate groups of rhythms are also indicated by the peculiar 
arrangement of time-bars. The intermezzo entitled " Paganini " 
in Schumann's Carnival affords an illustration. The measures are 
divided into four couplets : 


Op. 9, XVII. 

Ex. 328. - 

The upper theme (there is a counter-theme in the base) consists, 
of four eighths in each measure ; hence the notation. In the ninth 
measure and what follows there are only two groups, and, of course,, 
this indicates a different style of performance, thus : 

Ex. 329. 

The beginnings of each group are stationary tones, the melody 
notes being those which bear the accent marks. Compare this 
with Ex. 328. 

From the Romanza, op. 21, III, by Madam Schumann, an addi- 
tional illustration is taken : 

Ex. 330. 


Owing to the peculiar arrangement of the time-bars these motives 
are divided into groups of six notes each. The beginning of each 
group is to be slightly marked (especially since several motives in 
this opus commence upon the third eighth note), and then the ter- 
minating notes are to be somewhat disconnected from the following 
group. Otherwise a composer so intelligent and painstaking as 
was Madam Schumann would not have included both the bar divi- 
sions and the slurs. 

The rhythm of the accompaniment may be relied upon to pre- 
serve the mensural divisions, thus leaving the right hand free to 
express the melodic groups according to the notation signs : 


The last group in Ex. 330 merely shows another method of repre- 
sentation, which, in view of the slurs, is substantially the same : 
The connecting bars at (d) show that there is a motive group of 
six notes ; the slur shows the same thing, and indicates, further, that 
the six tones are to be connected together. The last tone at the 
end of each slur is therefore to be played lightly staccato. (See 
the Variations Brillante, by Chopin, op. 12.) 

Addition of the Poetic Text. In transcribing the Schubert 
songs Liszt frequently wrote the text of the poem in the piano 
score, thus showing the relationship between words and music in 
the original song. The idea is an excellent one, and might be used 
advantageously as an aid to artistic interpretation. For example, 
suppose The Wanderer is selected as a study. Follow the words 
as they appear in connection with the music, and endeavor to give 
the same general expression to their import which an accomplished 
vocalist would impart to the song. The poem inspired the com- 
poser ; why may it not inspire the performer also ? 

In similar manner the Serenade ; Hark, hark, the lark ; the Erl 
King ; and other Schubert-Liszt works should be studied. Like- 
wise the songs of Beethoven, Schumann, Franz, Wagner, Rubin- 



stein, Jensen, Saint-Saens, Massenet, Grieg, Lassen, Tschaikowsky, 
Franck, Lacome, would prove instructive if examined or transcribed 
in this synthetic manner. Such exercise would stimulate the imagi- 
nation and present a definite aim for the musical expression. More- 
over, it would reveal the particular significance which standard 
composers have intended to convey by means of certain harmonic, 
melodic, and rhythmic designs. 

A poetic sentiment thus defined (as in " The Dark Eye," by 
Franz) is decidedly preferable for the young pianist to the esthetic 
study of a purely instrumental work in which he must be governed 
by uncertain caprice or whim. This is particularly true in cases 
where the performer is not prepared to enter into communion with 
the composer. 

The startling realism which characterized the performances of 
Rubinstein and Stavenhagen in the Schubert-Liszt " Erl King," 
and the equally remarkable accomplishment of Miss Marie Brema 
in singing the original song, were directly inspired by Goethe's word- 
picture, the music of Schubert being the vehicle by means of 
which the effects were conveyed to the listener. 

Emotional Expression. Though the most important element 
of expression remains to be considered, this must necessarily re- 
ceive brief treatment. But with regard to punctuation, accent, 
phrasing, musical detail, etc., these are matters determinable 
through analysis. For this reason they have received extended 
notice as coming properly within the jurisdiction of theory. 

Increasing and retarding the movement (with the usual accom- 
panying dynamic effects) belong more particularly to the emo- 
tional style. A hastening of the beats corresponds to a quickening 
of the pulse, and represents some form or species of active emotion. 
(Beethoven, op. 57, 78, 81 (a) ; Chopin, A-flat Ballade ; Schumann, 
Etudes Symphoniques ; Brahms, finale B-flat minor concerto.) 

When we enter the realm of higher sentiment, passion, psycho- 
logic expression, we find, as in ocean depths and mountain caves, 
that the mysterious charms are hidden from view : all seems dark, 
impalpable, mystic. The diver* learns that even the unknown 
depths of ocean are illumined by electric agencies whereby the 
"tenants of the deep " clearly distinguish and guide their course. 


The owl and the bat perceive in the caverns of earth every crystal 
peristyle, every beauty-formed stalactite, not as we see them by 
an artificial light, but as they appear in their natural element, out- 
lined against apparent gloom by the glow of phosphorescent 

Whoever would reproduce the intended expression in such 
works as Beethoven's op. no, " Isolde's love-death," or Tschai- 
kowsky's op. 23, must experience a feeling analogous to that which 
the composer felt when he conceived the music. The performer 
must be influenced by the personality of the composer to such an 
extent that every nuance will seem a natural concomitant of the 
musical idea. 

Provided we are capable of experiencing the same emotions which 
moved the composer, our own personality must be lost in the realm 
of creative fancy. We must breathe a purer air, live better, hope 
stronger, aim higher, and, finally, exclaim with the poet : 

" Let me forsake the cold and crushing world, 
And hold communion with the dead." 



Style and Phrasing Influenced by Technical Considerations. 

Music of an elaborate character, especially in the accompaniment 
and passage work, is influenced to a considerable extent by the 
mechanism of the instrument for which it was primarily intended. 
For example, the following figure (a), could not be played on a 
violin, because the performer is obliged to reverse his bow for 
the descending group : 

Ex. 332. 

The style at (b) is well adapted to the instrument, and effective. 
A pianist might, however, easily perform the entire measure thus : 


For similar reasons a piano passage may be notated in a certain 
form which might otherwise seem whimsical. Such an instance is 
the following : 


Ex. 333. (a). 


"Papillons." N. Stcherbateheff. Op. 8. 

8va. . . 8va 

j8bjp= fd&ft==]=&=&f=t= 

'-##-> 0-\-^0+- -\--0-l\0-\- m -\ -f- -I g-f h- a 

^ i _^J_ f ^[ I | ^ 1^ | J 1^ I 9 \ I ^ [ I L I if_ I 




The twelve sixteenth notes are here divided into four groups on 
account of the manner in which they must be executed. After 
two measures more in this style the form of figuration changes, 
and then the rhythm falls into normal divisions of six sixteenth 
notes each. So far as note-values may be considered, the arrange- 
ment at (a) is the same as it would be if the right-hand part were 
written in regular groups, thus : 

A contrasting example is quoted from the principal theme of 
Chopin's nocturne, op. 37, II : 

Ex. 334. 

p: __i f: j __p _ 



A technical consideration undoubtedly influenced the composer 
in writing a double hooked note at -f- , for the interval is not 
unusual in modern piano figurations. Two violinists might per- 
form the entire phrase legato ; a pianist can not do so. 

The peculiar manner in which the slurs are drawn in the left- 
hand part also conforms to the character of the instrument and its 
technical management. A number of similar instances occur 
throughout the nocturne, which see. 

Indeed, these peculiar structural features of piano music have 
become somewhat idiomatic, and they enter largely into the distinc- 
tive literature of the instrument. One more example is quoted : 

Ex. 335. 

Moszkowski. Op. 36, V. 

The appendix contains a stationary tone as connecting link 
between the period ending in D and the one beginning in G minor 
at a tempo. In order to produce the intended effect the pianist must 
accent the d, sustain it firmly, and then perform the appendix in 
light pizzicato style, so that the vibrations of d will continue into 
the next measure. 

Among numerous other instances the following are cited : 
Jensen, Berceuse, op. 12, the accompaniment; Beethoven, first 
part of sonata, op. no ; Chopin, Nocturne, op. 48, II, second part. 

Designs : Melodic Harmonic Rhythmic. In order to com- 
prehend the general plan or the intermediate designs of the better 
class of music it is desirable that the student should carefully 
examine and analyze the work selected, away from the instrument. 
With the mind concentrated upon the score and unembarrassed by 
active performance, the pianist will acquire the art of theoretical 



analysis, and thus learn to observe many important details which 
are unnoticed by the average student. Aside from the inestimable 
advantage of such analyses to the student of interpretation, they 
reveal the structural features of a composition, and thus facilitate 
the process of learning the notes. For illustration an Album Leaf 
by Th. Kirchner is selected. The entire number is based upon a 
diatonic design, and soon as this fact is apprehended the performer 
may proceed, with the assurance that he already has a good under- 
standing of the music. But to the cursory observer the opening 
phrase would signify very little : 

Ex. 336. < 


Nevertheless, the design embraces a regular melodic progression 
in the base, the outline of which is this : 

Ex. 337. 

All the chords appear in their first inversion, and against this 
there is a stationary note above, which occurs in the second half of 
the first three phrases, like bell notes. When the cadence is added 
we have all there is (so far as structure goes) of the first period. 

In thus penetrating the design we simplify the task of memoriz- 
ing or otherwise mastering the composer's transcript, and at the 
same time our analysis throws more or less light upon the correct 
mode of performance. The second period also is simple in its 
construction. A sequence relation is observable throughout (as is 
usually the case with good music), and this it is which must first be 



apprehended. The first period presents, as we have seen, two 
designs : I. A uniform position of the accompanying chords. 2. 
A scale melody in the base. After examining the first phrase, 
and understanding the design, one ought to be able to play the 
two following phrases without the notes. 

The following simple illustration will serve as a guide to others : 

Ex. 338. 

A. Sandberger. Op. 2, IV. 

The lower part is plainly a counter-theme, and to be treated as 
such. Instances like this are innumerable. 

The Intermezzo in Schumann's Carnival contains a sub-theme 
in the base ; but this is in syncopation, and therefore not so ap- 
parent as are the other instances mentioned. It falls under the 

Hidden Melodies. An example from Rubinstein will more 
plainly illustrate this : 

Dance of the Bayaderes, I. 

Ex. 339. 






The melody here falls to the thumb and index finger of the right 
hand, and should be played somewhat in this manner : 

A hidden melody is more liable to be overlooked when it 
appears in connection with a harmonic passage. For instance, in 
Beethoven's op. 7 this motive appears in different voice-parts as 
leading theme during nine measures : 

Ex. 340. 

It occurs in the first movement after the second subject. A 
similar motive, similarly concealed, may be found in the coda of 
the master's op. 90. Instances of this kind occur in many modern 
works, especially those of Chopin and Schumann. 

Rhythmic designs may be divided into two classes: (i) With 
regard to the proportion of corresponding groups considered men- 
surally, or according to the actual number of notes in a model-group 
or figure ; (2) the peculiarity of certain motives with reference to 
the time-value of the notes. The first 1 motive in Beethoven's op. 

31, I, is an instance : 

|S I 

This peculiar rhythmic arrangement of notes plays an important 
part, especially in the development. Though independent of 
melodic movement, this rhythm is to be considered as motive when- 
ever it occurs, because it cannot be dissociated from the main theme. 

It is so in the rondo to the Waldstein sonata. The rhythmic 


is most persistently and 



artistically employed throughout the finale. Sometimes we hear 

only the characteristic Jf J ^ in different melodic situa- 

I > 
tions ; but in all these ramifications the rhythmic design is of first 

importance. And in the doppio movimento this is still more mani- 
fest : 

Ex. 341. 

fcfei - 

The arpeggio figures above are comparatively unimportant ; it is 
the rhythmic motive below which demands attention. Hence the 
author has included accent marks in the left-hand part, though 
these must be light, on account of the prevailing pianissimo.* 

Of rhythmic groups and figures a number have been quoted ; 
and, as we know, they exert a marked influence upon the style of 
performance. Such are the following, from d' Albert : 

Ex. 342. 

The groups (a) and (b) are rhythmically identical, though 
directly opposite in melodic movement. The last two groups (d) 
and (e) are variants of the preceding. A different form of rhythmic 
design appears above at (f), repeated in the same manner at (g) 

* See Rhythmic Imitation, chapter XI. 


and at (h). The style is sufficiently indicated ; if it were not, the 
performer would be justified in following the models at (a) and 

Modified Application of Dynamic Signs. The fact has al- 
ready been stated that marks of accent and emphasis are to be 
understood and applied relatively. In a romance or a song without 
words sf does not mean the same degree of force that it does in a 
tarantella or a stormy allegro. Nor does ff in the largo to Bee- 
thoven's op. 7 indicate as much tone as is demanded at the close 
of his op. 2, I, also marked^. 

Similarly, the dynamic sign p in Chopin's Berceuse is under- 
stood to indicate a softer quality of tone than is required in the 
military polonaise. The symbols -^^ ~^^==* likewise 

have a modified application in quiet movements, such as the 
Abendlied, by Schumann (from op. 85), or the Fable, by J. Raff. 
And in the Abendlied piano means something more than softly ; it 
also means dolce. 

In brief, every dynamic sign or symbol should be applied 
relatively, according to the character of the work under consid- 

Two Themes Combined. Principal Subjects in Juxtaposition. 
A distinction must here be observed between a regular melody 
accompanied by a counter-subject, and the combination of two 
principal themes in juxtaposition. In the former there is a melody 
and a consonant counterpoint proceeding simultaneously, but not 
equally, the main subject being predominant. When two prin- 
cipal themes are combined they usually carry with them an expres- 
sion of rivalry or contention, as of two opposing agencies. 

Instances of this kind will require a different style of perform- 
ance for each of the two motives ; these styles being influenced by 
the character of the themes thus combined. In orchestral music 
the great variety of tone colors at the composer's command en- 
ables him to set out the two themes in very marked contrast to 
each other. Thus, in the overture, " Sakuntala," the two motives 
are plainly distinguishable by means of contrast in the orches- 
tration : 



Ex. 343. 


& i- 


The trumpet motive presents such a strong contrast to the 
theme in the violins that the average listener readily comprehends 
both designs. 

The music of Chopin abounds in examples of this character, 
some of which require close analysis and discriminating treatment. 
The following have two or more themes in combination : Waltz, 
op. 64, III; Scherzo, op. 31 (the strain beginning in C-sharp 
minor) \ Nocturne, op. 15, II; Berceuse, op. 57 (first period); 
Etude, op. 25, III, and especially the duetto, op. 25, VII. 

A simple illustration occurs in the first part of A. Sandberger's 
op. 2, III : 



Ex. 344. 




Owing to the cantabile character of the tenor theme it requires 
a strict legato and considerable pressure touch. The isolated 
couplets in the soprano part will be sufficiently distinguished if 
they are given more lightly, because the higher tones naturally 

Such examples as occur in the coda to Chopin's D-flat nocturne 
are comparatively easy of management, since the rhythmic groups 
are alternate, not simultaneous. For similar reasons the sub-theme 
in the first of the cradle song requires the utmost care upon the part 
of the performer. Chopin's Etude, op. 25, VII, presents a very 
interesting example of two regular themes combined in duet form : 

Only the first measure of the upper melody is canonic ; this part 
is really an independent theme. The accompaniment falls to the 
middle parts, and these should attract very little attention. As a 
means of contrast between the two themes the upper one (being 
more cantabile) may be played legatissimo, especially the resolu- 
tions and such intervals as are marked with additional slurs. The 
more florid character of the lower theme brings it into sufficient 
prominence without employing undue accentuation. 

2 4 8 


The performer should endeavor to impress both themes upon 
the listener without sacrificing either the one or the other. 

Wagner employed the greatest number of leading motives and 
themes in combination. These appear even in the piano tran- 
scriptions by Liszt, von Biilow, Brassin, Bendel, Fique, and others ; 
but one must be familiar with the original scores in order to fully 
appreciate the significance of each separate theme. 


Single and Double Slurs. The recent modified application of the 
slur has resulted in considerable confusion among inexperienced per- 
formers as to the real significance of this important musical symbol. 

In vocal music the slur has two meanings : (i) To indicate a 
phrase or section to be sung with one breath (i. e., without a break in 
the tones) ; (2) to show that two or more notes are represented by one 
syllable. This indicates legatissimo, because if the singer does not 
blend the tones well together the syllable will be repronounced. 

In instrumental music the slur has various meanings : (i) To 
indicate legato, or the connection of tones Chopin Etude, op. 
25, II. (2) To distinguish the phrases which require punctuating : 

Ex. 346. 

Menuetto. L. Godowsky. 

-& L__-*-^ _*_0JIj P(^ ffj (S> f- 

a. ^--L * ^& m^-^r br L ^ * 

dolce. i 


(3) To make more manifest certain rhythmic groups by showing 
their separate relationship or their individual conformation. Thus, 
from the Chopin Etudes : 

Ex. 347. 

Op. 10, IX. 



If only one long slur had been used in each example it would 
indicate a style of performance slightly different from that of the 
separate slurs for each group. There was, however, no intention 
to indicate a disconnection after each group, but merely to indi- 
vidualize the groups ; and this it is which causes so much confusion 
in applying the slur. In the last example the separate slurs arc 
equivalent to a slight accent upon the beginning of each group. 
The apparent disconnection merely signifies non-legato. (4) Stac- 
cato passages, or notes with intervening rests, may be included 
within a slur (a bracket would be better) to show that they com- 
prise a motive or a phrase. In such instances the idea would be 
conveyed by means of accent, crescendo, diminuendo, or rallen- 
tando Clara Schumann, op. 10, I : 

Ex. 348. 

Scherzo I. 




IHS ipjl 


' T> 


*?? f 



J *f i *f 

I ' 9ffi 

9 Z 


- 1 1 J ; 

T , 


V L/ 


* Tr 









' 9 ~'" 
\ 1 

.1 !_.. 

-9- i 


The slurs (and the brackets) merely show the principal notes in- 
cluded in the semi-phrases. These are played as motives. The only 
legato effect is that of the resolutions, indicated by the short slurs 
below. In the next example, also from Madam Schumann, the 
application is similar : 


Op. 21, II. 

Ex. 349. < 

"S JV FTP- ) "I 

_. r 



This final cadence embraces the chief motive (in the tenor here), 
and this must be catenated, notwithstanding the pizzicato style. 

(5) To indicate disconnected couplets or other groups which must 
be separated one from the other, thus : 

Ex. 350. Chopin. Op. 10, III. 

r .......... u mmi^m ii U * 


* ! ^j^ i ~xj^~*^g~- ' I M*ff* ft : ! . I ~ 

At (a) the end of the slur indicates staccato, because the couplets 
are to be separated from 
one another. This read- 
confirmed bv the 

Ex. 351. 

mg is confirmed by 
peculiar notation signs. 
Therefore the interpreta- 
tion would be like this : 

At (b) the rhythmic groups are played legato as far as the slur 
extends ; then they are disconnected. The hooked notes at (c) and 
at (d) leave no doubt about this. A similar instance" occurs in the 
etude on black keys, op. 10, V : 

Ex. 352. 

The first note in each measure here is staccato, as is the first 
note on the second beat. The nature of the etude (brillante) 
justifies this style, and it also relieves the monotony which might 
otherwise result from those incessant triplet figures. 

The short slur is sometimes modified by a long slur, and in such 
instances the latter serves to counteract the tendency to disconnect 
the last note at the end of the short slurs. 

An illustration is selected from a prelude and fugue in E 
minor : 



Andante serioso. 


Ex. 353. 

33 5fe 

=^*-tr4 14^-*-* 
[*^j*- -w 

If this were played 
after the usual manner 
of legato couplets, 
thus : 

it would pervert the meaning of the music ; the style would be too 
light for the serious nature of this fugue subject. The chromatic 
progressions (resolving up and down a minor second) should, how- 
ever, be distinguished by connecting them more closely than the 
intervening skips. Therefore, a double slur becomes necessary : 

Ex. 354. 

A very slight accent accompanies the first note of each slurred 
couplet, and these resolutions are to be performed legatissimo. 
Meanwhile the long slur over the first phrase indicates unity and 
forbids the disconnecting of the couplets, as at (b). No actual 
isolation occurs until the d-sharp, at the termination of the long 
slur, is reached. Most of the short slurs in Chopin's Nocturne, 
op. 9, II, are to be understood in this sense, thus : 

Ex. 355. 

The upper slur (not included in the standard editions) removes 
all doubt as to the meaning of the syncopated couplets. In the 
same nocturne the couplets containing iterated notes are to be 
played in the usual manner, each couplet being disconnected : 



Ex. 356. 

The circumstances are, however, quite different here. 

The application of 
a modifying slur over 
small slurred groups 
is similar in violin 
music. Von Biilow Ex. 357. 
quotes a novel exam- (a). 
pie from Beethoven's 
op. 101, thus : 

In a marginal note by von Biilow we read : " The double slurring 
might mislead many readers. The explanation is simply this : 
g-sharp and f-sharp should be played with the usual legato ; 
f -sharp and e in a less con- 
nected manner. The fol- 
lowing notation could, there- 
fore, be adopted, but might 
also be misunderstood : 

while the original mode of writing, 
borrowed from stringed instruments, 
is familiar to all violinists " : 


See the adagio in Beethoven's op. 106, von Billow's edition; also 
Chopin etude, op. 10, X. 

Delayed Attack. Ritenuto in Forte Passages, etc. A 
short ritenuto or a momentary delay in attacking a climactic 
passage is frequently rendered necessary in order to give the 
requisite force to a full harmony. The former applies to a more 
or less extended passage wherein the harmonic structure is of a 
ponderous character. This is in keeping with the principle that 



large masses of sound, being less distinct, require more deliberation 
in their delivery than do smaller masses of sound. On this account 
the ritenuto is less noticeable than otherwise it would be. An 
example is quoted from a modern work. The principal theme 
occurs first as a base solo in unison ; then in the treble with accom- 
panying harmony. At the third appearance of the initial -period 
the theme is doubled and accompanied by heavy harmonic masses. 
Here the movement is taken less quickly : 

Ex. 358. 

Ph. Scharwenka. Op. 38, II. 

If this is not played in moderate movement, and very deliberately, 
it loses in decisiveness as well as in dignity, and fails to produce 
the effect intended. Also, the following repetition (ff) is to be 
taken poco ritenuto. 

A similar instance may be found in the Marche Brillante, by Raff, 
op. 132, where the main theme recurs for the last time. Likewise 
in the Andante from Tschaikowsky's 5th Symphony, transcribed by 
Richard Hoffman. 

With regard to the temporary pause or delay in attacking a forte 
passage, that must depend upon the situation in which it occurs. 

* The marks here signify that the tones must be sustained their full value. 



In the following instance we are perhaps influenced more by tech- 
nical than by purely musical considerations : 

Ex. 359. 

E. Nevin. Op. 13, V. 

td^y^sfc * 


+ F 



p 1 


j -^ 




... JP 



1 1 i A 

The triple dissonance marked ^=^ can not (by the average 
player) be struck simultaneously, but must be taken in arpeggio 
form. This will result in a slight detention of the melodic tone, 
especially if the discord receives its due amount of force. The 
next example is different in character : 

Ex. 360. 

Ad. M. Foerster. Op. 38, II. 

V [ 

It is not absolutely essential that the movement should be de- 
layed at ff t though a more decided effect is produced thereby, 
especially since we feel confident that the principal theme will recur 



at this point. Care must of course be exercised not to materially 
interfere with the tempo of the mazourka. 

Something similar occurs in the march at the close of Schu- 
mann's " Carnival." The expression is very determined and some- 
what pompous, and is played with the utmost deliberation, 
especially the groups marked ff and sf ' : 

Ex. 361. 

The marks, I, are not included as punctuations, but merely to 
show that a slight detention occurs in attacking the semi-phrases 
(a) and (b). The aim should be to give full force and energy to 
the first of each bracketed group, not to disconnect the tones, as 
in punctuating a phrase or section. This style of performance here 
gives to the music a peculiarly aggressive character, well suited to 
the composer's serio-comic idea in this finale : that is, the romantic 
element opposing the strictly classic. In other words, Paganini, 
Chopin, Liszt, and the Schumanns marching in battle array against 
Hummel, Diabelli, Spohr, Moscheles, Kalkbrenner, et al. 

Connected and Disconnected Ideas. Attention is here di- 
rected to another important feature revealed through analysis. 
The application is to dual subjects, intermezzi, the introduction of 
a new theme, or the appearance of an episode. A simple prelimi- 
nary illustration is selected from a rondo by Dussek. After part I 
has closed in G major the second theme appears, thus : 

*See the finale to " Etudes Symphoniques"; also " Marche Militaire," Schubert- 


Ex. 362. 

Meno vivo. 

Op. 21, I. 

The section beginning at (a) is a preliminary motive ; the princi- 
pal second theme commences at (b). The two ideas are quite dis- 
similar, and therefore it is not desirable to connect one with the 
other. A brief pause may be observed before beginning the main 
theme at (b). Also, the fact .is to be noted that the composer 
marked the preliminary motive/", and the regular theme/. This, 
together with the differences in key and in melodic construction, 
make it plain that the two ideas (a) and (b) are not to be connected 
in any manner, not even in movement. In such instances a differ- 
ence in the style of performance is naturally suggested. Com- 
posers usually separate or distinguish disconnected ideas by means 
of rests or pauses. Thus, in the Baracolle by G. Ehrlich part I 

closes in G major, and there is almost an entire measure t? rest 

before the beginning of part II in C, which is a contrasting motive. 
And in the G minor Air de Ballet by Moszkowski there is a fer- 
mata inserted after the unresolved /th chord and immediately be- 
fore the beginning of the finale in G major. 

In d' Albert's gavotte and" musette from op. i, a slight pause 
should be observed before commencing the musette. It is to be 
understood that the fermata does not apply until after the second 
period of the gavotte has been repeated, the object being to sepa- 
rate Part I from Part II, which is dissimilar. Chopin seldom in- 


troduced a second subject or episode for the mere sake of contrast 
(he had another method for attaining this end), and so there are few 
instances of this kind to be found in his music.* The causes 
which render this effect necessary are mostly of a dramatic or pic- 
torial nature ; hence we find numerous examples in the works of 
Beethoven, Von Weber, Liszt, Schumann, Saint-Saens, Tschaikow- 
sky, Sgambati, MacDowell, and Templeton Strong. 

Formal Outlines. The form of a composition, independent of 
detail, should be clearly defined by the performer according to out- 
line analysis. Thus the following structural features in Godard's 
" Au Matin" should first be observed from an examination of the 
score : It is a single form consisting of a short introduction, two 
consequent periods (one subject), and a coda. The first period is 
slightly extended ; the second period is united to the first. After 
these have been repeated they are varied, and then there is an ad 
libitum coda of 24 measures. Of course, these outlines can not be 
revealed at the outset, as in a picture, but if the opus is properly 
performed a good listener ought to be able to describe the form 
very nearly as it has been stated by the author. 

Another single form which requires careful analysis is Ad. Hen- 
selt's " If I Were a Bird." Every division and subdivision, as well 
as the contrasting tonalities, should be revealed by the performer, 
who must know whether the periods are curtailed, regular, ex- 
tended, or united. A clear idea of form is indispensable, but this 
can not be imparted here. 

Ad Libitum Repetitions. Repeated groups, especially in pas- 
sage work and cadenzas, sometimes require a farther continuance 
in order to produce a certain dynamic effect. For example, in the 
cadenza to Godard's B-flat mazourka the diminuendo (beginning 
from the lowest point) may be continued farther than the notes indi- 
cate in order to gradually reduce the tone quantity so that this 
effect will correspond to the gradual crescendo which precedes. 
The main object is to attain a properly graduated diminuendo' from 
forte to pianissimo. During the cadenza mensural accentuation 

* In the Prelude XV Mme. Carreno makes a slight pause before resuming part I at 
the close. 



disappears, and therefore no rhythmic impropriety will result from 
a few additional repetitions of the last group : 


When the tone has been sufficiently reduced, the opening figure 
(upon which the cadenza was constructed) is to be played a tempo, 
and with enough accent to indicate that the principal theme here 
recurs, thus : 

Ex. 364. 

Something of this nature occurs at the close of StcherbatchefT's 
divertissement in B : 

Op. 8, IX. 

Ex. 365. 


All the "signs and directions here are by the composer. " Bis 
ad lib." undoubtedly signifies that the player may continue the re- 
peated groups until the perdendosi effect is complete ; i. e. y until the 
sounds die away and vanish. 



At the close of Grieg's first " Peer Gynt " Suite there is a pecu- 
liar instance, directly opposite in effect to the examples quoted. 
After a violent forte passage the music suddenly ceases and there is 
a moment of silence. Then the kettle-drum begins a roll, at first 
very softly, but increasing in force until the extreme limit of cres- 
cendo is reached. Here the cymbals and the full orchestra make 
two strokes, like peals of adjacent thunder, and the music comes to 
an abrupt, almost infernal termination. Many will remember the 
electrical effect which Mr. Arthur Nikisch produced with this 
stretto. It is here indicated according to the author's recollection : 



Ex. 366. 

The pauses are not included in the original notation, but they are 
essential, and were undoubtedly intended by the composer. The 
tremolando in brackets includes about two measures, or whatever 
time may be required for working up to the fulminating climax at 
ff. Such effects can not be produced by the usual one- two- three- 
four method. The last measure is presto. 


Since the publishing of Bernard Boekelman's excellent edition 
of sixteen clavier fugues (selected from Vol. I and Vol. II of Bach's 
" Preludes and Fugues") the author's task is considerably simpli- 
fied. By means of notes differently colored, and in some instances 
differently shaped, the entire scheme and structure is plainly 
revealed. This edition is earnestly recommended, and the follow- 
ing brief dissertation is intended to supplement Mr. Boekelman's 
work. The essential ingredients of fugue construction are suffi- 
ciently indicated in the special edition referred to ; therefore it is 
presupposed that the more important features are at least tolerably 

The subject of a fugue is usually short and characteristic, and 
from this the entire movement is evolved. In fact, unity is the 
most eminent characteristic of a well-developed fugue. The 
answering voices proceed from tonic to dominant and vice versa, 
and it must be understood that the response (or " companion ") is 
also the subject. Therefore whenever subject or response appear,' 
either above or below, they are to be treated as principal theme. 
When the subject is concluded by the leading voice, it proceeds to 
counterpoint against the voice, which sings the response. If this 
counterpoint is used to accompany each entry of the theme (above 
as well as below), such accompaniment is called "counter-subject," 
and it is to be treated almost as prominently as is the main subject, 
because the counter-subject is a continuation of the motive. (See 
fugue in E-flat 7, II.) A distinction is to be made between a regular 
C. S. conceived in double counterpoint, and mere independent 
counterpoints. For instance, in the first fugue of Vol. I there is 
no clearly defined C. S., the subject and response being accompa- 
nied by freely invented melodic phrases. The same may be said 




of 8, I, and 2, II. However, the C. S. usually enters promi- 
nently into fugue composition, and serves as complemental accom- 
paniment to the theme, either above or below. The relative 
degrees of importance of the different parts may be thus stated : 
(i) The subject or response; (2) the C. S. ; (3) the additional 
counterpoints and ad libitum parts. The dynamic character of 
these various elements would therefore be in this proportion : 
Subject and response, forte ; counter-subject, mezzo-forte ; ad libi- 
tum or harmonic parts, mezzo-piano. 

These preconceived ideas are subject to modification under cer- 
tain circumstances. Where the rhythmic arrangement of the C. 
S. is in strong contrast to that of the theme, it is not necessary to 
observe any particular distinction between the tone quantity of 
subject and C. S. Such an instance is presented in the D-major 
fugue, 5, I. The subject is short and characteristic, and so differ- 
ent from the principal C. S. that the former requires no particular 
emphasis, except in the stretto. On the contrary, the sustained 
character of the main C. S. demands, special accentuation : 

Ex. 367. 

The three-voiced fugue in C-minor (2, I) is selected for purposes 
of illustration. This contains two counter-subjects, an important 
transition, and intermezzi in double and triple counterpoint. The 
fugue should at first be examined synthetically, observing closely 
the peculiar manner of its construction. For this purpose the 
Boekelman edition is especially valuable. The thematic develop- 
ment is remarkable, the closest affinity being preserved throughout. 
The fugue should then be analyzed in this manner : Play the 
principal theme (in red notes) separately as it occurs in the different 
voice-parts. Subject and response, being thematically identical, 
are to be phrased in the same manner. The theme occurs eight 
times, and in performing this separately the two hands should be 
employed alternately, or as in the complete performance. The 

FUGUE. 263 

first C. S. (in green notes) comes next. Play this separately, and a 
little more quietly, since it always serves as counterpoint to the 
theme. This is particularly necessary in such measures as 7,- 8-, 
15,- 1 6,- 26,- 27,- where the subject is below the C. S. The 
second C. S. (printed in violet ink) should then be extracted. This 
occurs in all the voice-parts. 

The transition and the intermezzi ("episodes") are printed in 
black ink, and there are also three kinds of characteristic notes to 
be observed. In the transition, or codetta (measures 5 and 6), an 
imitation of the leading motive appears thematically, accompanied 
by a contrary imitation of the first C. S. In the first intermezzo a 
fragment of the subject is treated canonically, with a paraphrase of 
the C. S. as lower counterpoint. The two upper voices are to be 
treated equally during the canon. The accompaniment in the base 
is also important. At 13 and 14 the second intermezzo appears 
inversely ; a fragment of the last half of the C. S. becomes the 
theme in two lower parts while the figure of the first part of the C. 
S. in reversed order serves as counterpoint above. The lower parts 
should predominate here. 

The third intermezzo, of three measures' duration, is more elabo- 
rate. The base has an imitation of the subject in sequence form, 
while the contralto counterpoints with the figure of the C. S A in 
contrary movement. Base and contralto then exchange parts (in- 
versa). The circular notes O i n each instance indicate the theme. 
The soprano part is ad libitum, and, therefore, to be played more 
softly. These constitute the principal features, and after they have 
been thus analyzed the fugue may be attempted in its entirety. 
Subject and response are phrased in the same manner see measures 
I and 2. During the progress of the fugue, especially where the 
subject appears below, this uniform style for each part will aid in 
distinguishing the main theme from its C. S. A few additional 
suggestions are included : 

Ex. 368. 



9. 10. 



These styles and groupings are easily comprehended, and the 
technical requirements are not great. But to play the fugue well 
requires considerable skill and intelligence. Especially where the 
subject or response falls to the middle voice-part (as in measures 
1 5-16) it is difficult to apply the relative degrees of tone quantity to 
the various themes and counter-themes. When the subject first 
appears in the base (7 and 8) it is comparatively easy for the 1. h. 
to give the necessary prominence to that part; but such instances 
as occur at 15 and 16 are not so easily managed. 

And finally the more important problem presents itself: How 
preserve the strict polyphonic character of the fugue, and clearly 
indicate all the entrances and exits of the theme, without causing 
the composition to sound like an exercise in counterpoint ? 

Pn the other hand, there is danger of committing a still greater 
error by infusing into a fugal work the extravagant sentimentality 
of our age of which good father Bach was happily ignorant. 

It is a noteworthy fact that nearly all the fugues by Bach are 
constructed differently, and this adds to the difficulty of their per- 
formance. Sometimes the theme is treated inversely, as in 6, I ; 
and these are equally important. At other times the subject 
appears in augmentation, 2, II. In some of the fugues there is no 
C. S. ; in others there are two or three, and frequently there is but 
little regular "exposition." This leads to the remark that too 
much stress is usually laid upon the text-book formula of fugue 
construction. From this the student is inclined to underestimate 
the value of the so-called " episodes," though in the fugue under 
notice these intermezzi occupy about the same number of measures 
as do the expositions. In 3, I, there are less than 24 measures 
of exposition^ whereas the development will aggregate 31 measures. 


26 5 

15, I, is another instance in which the intermezzi and developments 
are really more important than the merely formal exposition of 
subject and response in their prescribed order. 

Fugue is the most compact and coherent of all music forms. 
Adventitious parts seldom appear, and since every voice has its 
own individual melody, the task of the performer is much more 
difficult than it is in lyric or harmonic music. Pianists as well as 
organists might benefit greatly from the example of M. Alexandre 
Guilmant, whose fugue playing was one of the greatest revelations 
in the author's experience. An ideal fugue performance requires a 
separate voice or instrument for each melodic voice-part, and these 
requirements should be borne in mind by all solo performers. M. 
Guilmant's remarkable method and interpretation very nearly ful- 
filled these ideal conditions. 

In addition to Bach's "Well-tempered Clavichord" and the 
" Five Fugues " by Haendel, the following are mentioned : Three 
Preludes and Fugues, Clara Schumann, op. 16 ; Piano fugue in 
G-minor, Rheinberger ; ditto, Guilmant-Haberbier (Fugue in D 
by Guilmant, transcribed for piano by Mme. Rive-King, the pre- 
lude by Haberbier) ; MacDowell, Prelude and fugue, op. 13 ; and, 
finally, the fugue-fantasia in Beethoven's op. 106. 


Variety of tone color has become such an essential element of 
high-grade piano playing as to demand a systemized mode of 
procedure as to how and when special effects may be produced. 
Suggestive imitations of orchestral and other instruments naturally 
form the basis of this attempt to outline such a method. And 
even though one may seldom have occasion to imitate orchestral 
tone color, yet the ability to do so will prove of inestimable value 
in performing a wide range of piano literature. The mere sugges- 
tion of a definite tone quality at a certain point in an opus often 
stimulates the imagination of the performer by presenting to his 
mind a tangible ideal. 

Furthermore, there are numerous instances in which it is not 
only legitimate, but essential, to approximate as nearly as may be 
the peculiar timbre of certain orchestral instruments. Mr. Arthur 
Friedheim's performance of the " Tannhauser " overture is here 
cited, because of the pianist's successful attempt to reproduce par- 
ticular tone colors according to Wagner's original score. Also 
the March Brillante from Raff's suite, op. 91, as played by William 
H. Sherwood. The author wrote on his program at the time, 
" wood wind," the suggestive imitation having been remarkably 

Let us suppose that an inexperienced player undertakes such a 
work as the favorite march from Raff's " Lenore " symphony. 
The performance is colorless and uninteresting. Now, suppose, 
farther, that the principal indications in the original score are ex- 
plained to the performer. The effect will be greatly enhanced, 
even though the interpretation still remains unsatisfactory. The 
prelude is here quoted, with the necessary indications as to timbre : 



26 7 



Ex. 369. 



These effects are easily produced if one knows how they sound 
from kettledrums and horns, according to the original score ; and 
certainly a clearer impression is conveyed to the listener by means 
of these suggestive imitations. Even in the piano sonata by 
Beethoven, op. 53, von Biilow, in his special edition, wrote over 
certain phrases, quasi oboe, quasi flauto, quasi fagotto, etc., partly 
because the great composer usually had an eye to the orchestra, 
but principally as an inducement to the pianist to change the tone 
quality during the antiphonal motives. (See von Billow's explana- 
tory foot-note in reference to these imitations.) 

Accompaniments to concerti, when played on a piano, demand 
considerable skill in this respect, since the character of the work is 
frequently sacrificed by an achromatic accompanist. An extract 
from the coda to the first movement of Beethoven's op. 37 is pre- 
sented as an illustration : 

Ex. 370. 



i- r i- ' -( ( 

-^ N- 


C ' ' i- 

3? . IT.^ 3 

*The concerto is founded upon this timpani motive. 



The effect of strings and timpani answered by the arabesk figures 
on the piano is very charming, and certainly there is no good rea- 
son why this effect should be wholly sacrificed because an 
orchestra may not be available. The design here is very favorable 
to these imitations on the accompanying piano, because the base is 
a regular timpani motive and the kettledrums are more clearly 
suggested by a light and almost indistinct quality of tone. The 
accompanist must bear in mind that in solos of this kind the vibra- 
tory quality of the kettledrums is considerable. Where the tone 
is of very brief duration, it is because the timpanist damps the 
sound with his hand. 

With regard to the French horns, their tone is clear and ex- 
tremely musical. The damper pedal aids somewhat in approxi- 
mating their effect. The trumpets are piercing in quality. With 
regular trumpet motives this instrument can plainly be suggested. 
The trombones are naturally blatant in forte passages, though the 
tone is majestic and noble. In the hands of an artist it can be 
sustained very softly. Wagner has written pianissimo chords for 
3 and 4 trombones which sound like the organ in a solemn 
cathedral service. (The evening star romance from " Tannhauser " 
is an instance.) Effects of this character are necessarily more 
difficult for the pianist to attain, but the peculiar strident tone of the 
trombones in harsh passages is easily reproduced. 

The almost total absence of overtones from the flute serves to 
render its tones clear and limpid. In aiming at this quality of tone, 



wherein clarity is the chief peculiarity, a very distinct, quasi staccato 
touch should be employed. 

The violins are not so easily suggested. Even the greatest 
organ builders have failed to reproduce the effect of bow and 
string. There are, however, certain passages wherein an approxi- 
mate representation of the violins may be attempted. This pre- 
supposes that the passage is antiphonal in style and that a natural 
violin figure forms the basis of our attempted imitation. The 
pizzicato is easily managed by means of a light and short finger 
staccato. Also the violin tremolando may be fairly represented on 
the piano, using the damper pedal and making the reiterations rapid 
and non-pulsatalic. 

The ordinary harp effect is somewhat similar to that of the 
violin pizzicato, but the former is not so short. To imitate the 
plucking of strings it is necessary to employ a quasi staccato 
touch, with the dampers open. See " L'Arpa," by Raff, op. 17. 

Bell tones have been described elsewhere in this volume. They 
can be very well represented by the pianist, and frequently demand 

Guitar, mandolin, banjo, bagpipe, music box, all can be imitated 
should occasion require any of these peculiar effects. 

A brief antiphonal passage is quoted from the " Oxford " sym- 
phony : 

Ex.371, (a). 


This is from part II of the minuet. The only difficulty consists 
in representing the pizzicato of the strings against the syncopated 
melody of horns and bassoons. The latter requires open dampers, 



whereas the pizzicato must be played with closed dampers. One 
may therefore sacrifice the pizzicato in favor of the more important 
horn and bassoon effect ; or, by managing the damper pedal in 
this manner the complete representation may be attempted : 


S T*r*x*z^fe^^*= 


The pizzicato, light and detached, should mark the measures. 
The antiphonal phrase by strings is to be played smoothly and 
with as little accent as possible. 

Effects similar to those at the end of the Swiss melody, in the 
" William Tell " overture, are frequently met with. As the last 
tone of the Alpine theme is sounded from the oboe the trumpet 
rings out an alarum, which is immediately developed into a bustling 
fanfare : 

Ex. 372. 



= = = 

- 1 i i - i [_ ii_(_ JH =a i 

--U- I O ~ 

J_ --- -^^ 
i^EEEEEi: =F F- =3 

The accent (which is here very marked) applies to the trumpet 
note only ; all the other parts end softly on the united period. 
In performing transcriptions of overtures, symphonies, and other 



orchestral works, the pianist should be familiar with the original 
instrumentation in order to give at least a suggestion of the more 
characteristic tone-color effects. 

And even with music written expressly for piano it is evident that 
a slight modifying of these various tonal qualities may be applied 
in expressing the manifold sentiments with which modern music 

At the close of the next chapter the author has enumerated a 
number of instances in the Beethoven piano sonatas wherein certain 
orchestral tone colors may be suggested. Whether these sugges- 
tions are finally adopted or abandoned, the fact remains that such 
tentative efforts are very beneficial to the young pianist. 

*See the intermezzo in Chopin's military polonaise fagots, string bases, and drums 
answered by trumpets and trombones, 



No one who reads musical history intelligently can doubt that 
different ages possessed certain peculiarities of style and expression. 
These were influenced by the conditions of the art at a given 
period, by religious thought and impulse, by literary and political 
revolutions, and by the nature and quality of musical instruments. 

I. The Sixteenth Century. Until the latter part of the six- 
teenth century the art of music was confined principally to vocal 
works, as exemplified in the church compositions of Palestrina, 
Tallis, Lasso, and other masters. Their music expressed the 
religious spirit of the age : classic in outline, formal in construction, 
and intolerant of innovation. The canons of musical art were 
rigorously enforced, and the all-pervading influence of dogmatism 
and asceticism hampered the productive efforts of composers and 
militated against the charm of natural expression. An important 
clue to the interpretation of those works is thus furnished, and 
it leads beyond the melodic and dynamic indications of the score. 

Instrumental music was a later development. The organ com- 
positions of Frescobaldi may be mentioned first. Though counter- 
point, in its scientific aspect, had been highly developed, harmony 
was but imperfectly understood. Also, the organ was primitive in 
design and crude in construction. But despite these disadvantages 
Frescobaldi was a great pioneer virtuoso and produced a consider- 
able number of important works for the organ. The principal 
characteristics of the vocal music of that time entered into the 
instrumental works of Frescobaldi and his immediate disciples. 
Formality and austerity are the dominant features of their music, 
and these traits should largely influence the manner of perform- 
ance. This epoch includes Reinkin, Frohberger, and Buxtehude, 




though these masters undoubtedly improved upon the original 

II. Early Chamber Music. During the seventeenth century the 
art of viol making was more highly developed than that of any 
other instrument ; consequently, the early violinists possessed great 
advantages in this important respect. Torelli and Corelli un- 
doubtedly commanded excellent violins, and their compositions 
naturally included more euphonious melody than is to be found in 
the organ music and masses of that period. The technical devel- 
opment of the violins was limited, but some of the music, particu- 
larly that of Corelli, is happily conceived, melodious, and skilful in 

The chamber music of Lock, Lully, Purcell, A. Scarlatti, and 
Couperin may be included in this epoch. 

III. The Harpsichord and Clavichord Epoch. A more impor- 
tant period began with Rameau, Purcell, D. Scarlatti, Haendel, and 
J. S. Bach, and it includes Tartini, Sammartini, Paradisi, Galuppi, 
and P. E. Bach. The music of this epoch is mostly thematic, this 
style having been influenced by the spinet, clavichord, and harpsi- 
chord then in vogue. Canonic imitations pervaded almost every 
form. While this style admits more artificiality in its construction, 
it has the advantage of more logical development than appears in 
the lyric style, which followed. One motive was sufficient for a 
movement in thematic style, whereas there are eight different periods 
in Haydn's Gipsy Rondo. Such pieces as the Lesson in G by 
Haendel (quoted in chapter XIX), Toccata in A by Paradisi, or the 
" Solfeggio" by P. E. Bach, illustrate the peculiarities of the 
Couperin-Bach epoch. The music is direct, consistent, and whole- 
some, and it admits but little deviation from the prescribed move- 
ment and the prevailing mood. Grace and charm of expression, 
as well as vigor and incitement, are demanded of the performer ; ! 
but he must ignore the present and live for a time in the past. He 
must woo the spirit of a bygone age, in which steamboats, rail- 
roads, and electric telegraphy were unknown. Alas ! there are few 
now who will suppress the feverish tendencies of our age, even for 
the sake of communing with the exalted spirit of good father 




The author does not intend to imply that the lyric element was 
unknown during this epoch, for it was considerably developed by 
Carissimi, the elder Scarlatti, Purcell, and all the opera composers 
of that' time. But the keyed string instruments, from the time of 
Frescobaldi until after the birth of Mozart, possessed so little sus- 
taining power as to discourage lyric composition. Hence the 
prevailing thematic style, with its countless and curious agrements, 
fiorilure, and manieren. 

A tonal peculiarity of this epoch should be mentioned. The 
passing notes and appoggiaturas were almost invariably diatonic, 
seldom chromatic. Whereas in music of a later date the ascend- 
ing appoggiaturas and certain passing notes are written a minor 
second below the harmonic or principal note. The difference 
occurs when the unrelated note is below the principal note, thus : 

Old style. 

Ex. 373. 

In translating the signs and symbols of music composed during 
the fifteenth, sixteenth, and seventeenth centuries it is necessary to 
understand these tonal distinctions, because they really do exist, 
and should therefore be observed in performing the grupetto, the 
mordent, and other ornaments indicated by means of symbols. 

Even as modern a composer as Dr. Callcott employs the diatonic, 
in preference to the chromatic method, as here : 

Ex. 374. 



Attention is directed to the base solo. The diatonic passing 
notes in measures one and two are not only more quaint, but more 
expressive of the sentiment (" Once upon my cheek he said the 
roses grew ") than the modern chromatic method would have been. 

These tonal peculiarities are observable in the harmonic as well 
as in the melodic construction of the music under notice. The 
result was a freer use of secondary (non-transitional) seventh 
chords than in the music of Haydn, Clementi, and Mozart. The 
older method has this advantage, that it presents a greater variety 
of harmonies ; the modern method employs the principal seventh 
chords more frequently, and, therefore, while the harmonies are 
less varied in character, the changing tonalities offer a greater 
variety of keys. The former is illustrated in this example : 

Ex.375, (a). 




There are here four species of seventh chords, indicated by Roman 
numerals according to the author's Analytical Harmony. These 
secondary discords are much more reposeful than a series of princi- 
pal seventh chords would be. In fact, the almost incessant use of 
transition chords is one of the signs of our impatient, electrical age. 

In the last example a dominant relation is maintained in the 
sequence, and yet the tonality 
of G is not disturbed by even 
a passing modulation. 

The modern tendency 
would be to change the sec- 
ondary into principal discords, ^TL* ^ n*. m. 

thus substituting the chrom- 
atic for the diatonic element : 





Occasional instances of this kind maybe found in the harpsi- 
chord works, but the prevailing tendency was toward diatonic 
progression. The influence of this upon style and interpretation 
is so manifest that the author deems it unnecessary to offer further 

Though Bach was far in advance of his time he adhered very 
closely to the polyphonic style. Lyric themes may be found 
among his works, but these are usually accompanied contrapuntally 
rather than harmonically. His song, " My Heart Ever Faithful," 
is an instance. Nearly all his clavier music is instrumental, rarely 
vocal in style, the themes being motivized as in fugue construction. 
It is so with the clavier music of Couperin, Purcell, Haendel, and 
other composers of that epoch it is mostly thematic, rarely 
lyric. These styles having been explained and illustrated under 
their several heads, we may pass on to 

IV. The Mozart Epoch. This was a lyric age. The clavichord 
and harpsichord began to be superseded about 1750 by the forte- 
piano, which possessed greater volume and a more sustained tone. 
Chord progression, in its application to harmonic accompaniment, 
also became more diversified and more generally understood. It is 
unnecessary here to exactly trace the origin of harmonized melody 
as it so abundantly appeared during this period. But the improve- 
ments in keyed string instruments undoubtedly enabled Boccherini, 
Haydn, Clementi, Mozart, and other composers to treat the forte- 
piano in a more lyrical manner. The new style seems almost to 
have leaped into existence and favor. For example, compare the 
A-major toccata by Paradisi (1710) with the finale of a sonata by 
Haydn (1732), or the Solfeggio by P. E. Bach (1714) with the 
Mozart fantasia or rondo in D. The pieces by Haydn and Mozart 
will seem to belong to a different age, though all four composers 
were living in the year 1770 ! 

In addition to the more tuneful element of the new instrumental 
lyric style, there is also less strictness of movement than was ob- 
served by the earlier clavier composers. And on account of the 
greater importance attaching to melody, it may be stated that a. 
more strict legato is required for the music of Mozart than for that 
of Scarlatti. The style and expression became more human 



and .less scientific ; more musical, but less logical. (It is 
easier to analyze a rondo by Mozart than an allegro by Bach.) 
From contrapuntal theorem and rhythmic contrivance we pass to 
the almost artless simplicity and naivete of folk music. A greater 
contrast can scarcely be imagined. 

We read pretty rhapsodies about the "pathos " and " tender sen- 
timent " of the old clavier music, but it rarely sought to express 
more than the artistic illustration of theory. There is design, as in 
tapestry or a mosaic ; symmetry, as in architecture, and logical ap- 
plication of principles, as in all artistic accomplishment. Major and 
minor were symbols of sunshine and shadow ; rhythm represented 
the impelling force in nature. For the remainder, it was a theoreti- 
cal problem : How to build a musical structure in a certain form 
and create variety without sacrificing unity ? It was a test of the 
material to be employed in composition, as here : 

Ex. 376. 


J. E. Bach. 




A two-part design in suspension serves as accompaniment to the 
theme below. It forms good counterpoint and is consistent, but 
does it signify anything further ? Does it express a scene, a senti- 
ment, or an emotion ? Assuredly not. We should, however, ren- 
der due homage to those sturdy pioneers who broke through forest 
gloom a path which led Mozart and his followers to the sunlit ely- 
sium beyond. Without Monteverdi, Scarlatti, Couperin, and Bach 
there could have been no Mozart or Beethoven, no Chopin, Tschai- 
kowsky nor Wagner. 

This era, beginning about 1760, includes the first period of 
Beethoven and a considerable portion of the music of Mehul, Schu- 
bert, Von Weber, and Mendelssohn. It dealt almost exclusively 

2 7 8 


with human affairs. Man comes upon the stage and plays his part. 
There is sorrow and joy, tragedy and burlesque, domestic scenes 
and rustic life, emotional excitement, vain striving, and various 
other moods and sentiments with which we are familiar. 

A noticeable feature of the music of this Mozart epoch is the 
regularity of its period construction. The lyric themes were usu- 
ally periodized in a uniform manner, the rhythmic phrases being 
equal, as in popular rhymes. Thematic works we have found to 
be much less regular in this respect, the periods usually being un- 
even or extended. 

In Chapter XX the author has attempted to show that lyric 
compositions require very little punctuating or special accent in 
order to reveal the outline divisions, which are usually well marked 
by the regular organic structure of the melody, thus : 


Ex. 377. 

The performer may bestow ever so much care upon the delivery 
of this simple theme, but if he undertakes to " phrase " it (as many 
do), he will most assuredly destroy that simplicity of expression 
which is its principal charm. In the great majority of such in- 
stances the usual "method of phrasing" is a palpable absurdity. 

The dominant trait in Mozart's music is tenderness. Simplicity 
and cheerfulness, directness of purpose, regretful yearning, and 
occasional heroic moments, find expression in the violin and piano 
works, the quartets, quintets, and symphonies. But the smiles and 
the tears are nearly akin. Too gentle to scorn the sordid world, 
and meeting but little substantial encouragement, Mozart resigned 
himself to whatever fate might decree, and he was thus induced 
frequently to fulfil the composer's task in a careless or perfunctory 
manner. He required a special motive to stir the calmness of his 
soul life. The " Jupiter " and the last G-minor symphony show 
the depth and versatility of his genius when an incentive did appear. 
This is still more apparent in the operas and the immortal swan 



song. After listening to these it seems incredible that such a com- 
poser could have written the Turkish March in his A-minor 
sonata ! 

Having mastered the style of Mozart, it will be an easy matter 
to interpret his contemporaries. In form and outline the sonatas 
of Clementi and Hummel are similar, but there is more formality 
and less poetry in the music. 

Dussek was a follower of Mozart ; so was Steibelt. Even 
Beethoven worshiped at the Mozart shrine, and for a considerable 
period we can trace distinct echoes of the Salzburg master. But 
ere long there came a new dispensation, and the Flemish tone poet 
was"its prophet. 

V. Beethoven. The principal clue to Beethoven's music is to 
be derived from a knowledge of the man and the artist. He was 
an emancipator, a poet, and a philosopher ; a herald of futurity. 
We must follow his varied and storm-crossed career, weep for his 
suffering, rejoice in his moments of victory, laugh with his merri- 
ment, penetrate his motives, and stand against the world for the 
psychologic art creed of the choral symphony. 

Almost the entire gamut of emotional expression is to be 
sounded, and every possible variety and shade of tone color is de- 
manded of the recreative artist. 

Orchestral effects occur in many of the solo sonatas. For ex- 
ample, op. 2, III, first allegro : 



Ex. 378. 

After this there is a cadenza, violins ascending, then two flutes. 
And in the following adagio this dramatic effect : 




Ex. 379. 

Also the largo, op. 7, measure 20, after the deceptive cadence : 
horns, trumpets, and trombones. Measures 37-38 : horns, bas- 
soons, and double basses, then flutes. Op. 10, II, Part II : horns 
and strings. Op. 10, III, the menuetto : woodwind. Op. 22, the 
principal motive : 


Ex. 380. 

)ET - j ^ _; ^ 

Op. 27, I, last allegro : brass, woodwind, and strings in antiphonal 
semiphrases. Same movement : trombones responding to the 
trumpet : 

Ex. 381. 

te EE$EE 

Also op. 28, Part II of the andante : horns and bassoons answered 
by clarinet and oboe, or woodwind answered by strings. Op. 31, 
II, the opening : harp and horn, then the violins. Adagio, same 
opus, measures 17 to 22 : timpani, violoncelli, and fagotti. 

These examples were selected almost at random from Volume I 
of the piano sonatas, and are merely intended as hints to the 
young pianist. 


VI. The Nineteenth Century. The epoch of Chopin and 
Schumann is the most important in piano literature, especially since 
it includes the greatest of all keyboard virtuosi, Franz Liszt. 
When Kalkbrenner remarked that he would like to give Chopin 
some piano lessons, because " his method was very faulty," the 
'famous professor merely voiced a belief generally entertained at 
that time that the music and the performance^ of Chopin alike 
were heterodox. And so they were. 

The works of Chopin and Schumann do not, as commonly 
supposed, represent anti-classicism, but art development. Piano 
manufacture had made rapid progress ; the material of composition 
had been greatly enlarged, and Chopin and Schumann merely ex- 
pressed in their individual ways the spirit of the age in which they 
lived. The music changed with changed conditions, and since the 
tonal expression was new, so was the style of performance. Greater 
variety of harmony and rhythm, more sparkling brilliance, less 
conventionality, finer nuances, and withal a certain mysterious sig- 
nificance (which the interpretative artist must discover for himself) 
these are the principal characteristics of Chopin's and Schu- 
mann's best works. 

Alas ! what vain attempts are made to cajole and conjure the 
spirits of those ever-living masters ! Loud and soft, fast and slow, 
these are the weapons with which amateurs and alleged pianists 
attack and mutilate the creations of romanticism. 

It is true that the best music of this period admits greater free- 
dom of movement in performance ; but it is not true that all sem- 
blance of regular tempo disappears. Chopin, Schumann, Liszt, 
Rubinstein, all have recorded their opinions on this point, and the 
unanimity which characterizes their various statements leaves no 
room for doubt as to the proper application of that much-abused 
license, tempo rubato. It is recorded of Liszt that while hearing 
a lesson he severely criticised a pianist for his spasmodic acceler- 
ando and ritardando. In answer to the master's query, " Why do 
you so ? " the pupil gave as a reason that the music was marked 
" tempo rubato." " But," replied Liszt, " that is not tempo 
rubato. Come here by the window and I will show you." 
Pointing to a large shade tree hard by, he continued : " You ob- 



serve the swaying of those branches and the agitation of those 
fluttering leaves, but the trunk of the tree remains firm and steady. 
That is an illustration of tempo rubato." 

It is known that the peculiar expressiveness of Chopin's playing 
was in large measure owing to the ad libitum style of the right- 
hand part, while the left hand maintained the regular movement of 
the accompaniment. 

Melody notes may thus be shortened or lengthened, declaimed 
or sung, without arresting the progress of the music or disturbing 
the poise of the accompanying background. Brief examples of 
this have been given in Chapters XIII, XX, and XXVI. 

Slow movements are more susceptible to these effects than are 
fast movements. A quick rate of speed is a sign of motion and 
animation, whereas a slow movement is associated with meditation, 
deliberation, and repose. Moreover, the expressive deviations from 
the regular beats of an andante or largo do not suggest that un- 
pleasant, hysterical effect which results from an unsteady allegro 
movement. It was this uncertain " drunken gait" against which 
Schumann so vigorously protested. The tendency of the present 
time is to exaggerate musical expression into bathos and to de- 
grade psychologic emotion to the level of mawkish affectation. 
Yet the example of all great pianists is against this misapplication 
of tempo rubato. Not alone the profoundly analytical Von Biilow, 
but the electric Rubinstein, disdained this convulsive style of per- 

The works of Schumann are even more romantic and mystical 
than those of his great contemporary. If the German was less 
poetic, he was also less mutable. There are, however, many fea- 
tures of style common to both composers, though each must be 
studied independently and sympathetically. Schumann was one of 
the most unfortunate of men, but his soul-life was as beautiful as 
any which history records. All this, and more, is impressed upon 
his musical creations. 

' The changes which have taken place since the death of Chopin 
and Schumann are unimportant as far as piano literature is con- 

The various styles of the present period present insurmountable 


28 3 

obstacles to the interpretative artist. This is an eclectic as well as 
an electric age. It does not seem possible for a violinist or a 
pianist to faithfully represent or reproduce all the styles which have 
been described, because some of these conflict with others. From 
the stupendous psychologic tone impressions of a Tschaikowsky to 
the thematic sphinxes of the sixteenth and seventeenth centuries is 
too great a span to be encompassed by a single artist. Rubinstein, 
with his almost universal genius, came the nearest to a solution of 
this many-sided problem. Yet even he could not wholly satisfy 
the demands of every musical epoch. 

Unfortunately, there is too much of eclecticism and convention- 
ality in the arranging of our recital programs. Would it not be 
better if artists confined themselves to compositions with which 
they are in touch and sympathy? Temperament rather than 
custom should be the controlling influence in the choosing of solo 


Roman numerals refer to chapters; cardinal nutn&ers, to pages or examples. 



Accessory parts. See Accompaniment, 

Acoustics, XXI, XXII, XXIII. 

Ad libitum. See Antiphonal Phrases, 
Parenthesis, Cadenza, Coda, Recitative, 
and Recollection, XI, XII, XIII ; also 

Agremens, XVI ; also the Harpsichord 
and Clavichord epoch, XXX, p. 273. 

Anticipation, XI, Exs. 106, 107. 

Antiphonal style, XI, XXIX, XXX, Bee- 
thoven epoch. 

Appendix codetta, Ex. 130. 

Avanera. See Habanera, XV. 


BASE basis, XIII, XXIII, Exs. 280, 287, 

288, 289, 290. 

Basso ostinato. See Ground-base. 
Bell motives. See Carillon. 

Continued thesis, XII and Ex. 134. 
Counter- theme (C S.), XI, Exs. 117, 119. 


DANCE rhythms, XIV, XV. 
Dissonance, XXII. 
Drone-base, Exs. 143, 150. 
Dual themes, XXVI. 



Eingang, XII. 

Embellishments, XVI. 

Emotional expression, XXV. 

Episode, XII. 

Extended periods, X, XIX. 


FERMATA, V, VI ; also disconnected mo- 
tives, p. 256. 

Fioriture, XVI. 

Form formal outlines, XIV, XV, XVIII, 

Fugal style, XXVIII. 



Cadenza, p. 97. 

Canon canonic imitations, XI, XIX. 

Carillon, XIII and Ex. 336. 

Coda, XIII. 


Gruppetti, XVI. 


HARMONIC accompaniment, XXIV. 
Harmonic form, XXI. 





IMITATION, XI and Ex. 225. 

Intermezzo, XII. 

Introduction. See Prelude, XI. 

Refrain, XII. 
Rhythm, XI, XIV, XV, XVII, XXVI. 
Ritardando, V, VI, VIII, XXVIII. 
Ritenuto, XX ; also Recollection, p. 123. 

Inversion. See Counter- subject ; also Ex. Rubato, XVII, XXV, Ex. 193 (a) and (b). 

L- S. 

LEGATO legatissimo, III, IV, V, VI, SEQUENCE, pp. 83, 84. 




Signs and symbols, XVI, XXV. 


Stretto, XIII. 

Style, XIX, XX, XXI, XXV, XXX. 

Monotone, Exs. 195, 260, 261, 336, 341. Sub . theme) Exs< II7> Il8 (b)> I95> 247> 

Mordente, XVI. 


Nuance, XVI. 


Organ-point. See Pedal-note, XIII. 
Ornaments, XVI. 


Passage, XII. 
Pedal-note, XIII. 

306, 307, 309. 


Timbre. Tone-color. 
Tonal considerations, XXII, XXIII ; also 

movements influenced by key, p. 1 68. 
Tone-color, XXIX ; also Exs. 378 to 381. 
Turn grupetto, XVI. 


UNEVEN periods, IX, X, XIX. 
Piano pedals, Exs. 151, 152, 257, 259, Uneven phrases, VII, XV (czardas). 

287, 288, 289, 290. United periods, X. 

Polyphony, XXVIII. 


QUINTOLE, pp. 147, 148. 


Recollection, XIII. 


VOICE- PARTS, Exs. 109, no, 1 1 6, 117, 
206, 248, 278 ; also XXVIII. 


ZARABANDA. See Sarabande, XIV. 

Abbreviations for Marking the Analyses, 
with Corresponding Key. 

NOTE. Elementary students may mark the phrases with a comma, and the end of a 
period with O. They should also indicate whether the periods are regular, curtailed, 
united, or extended. Then they may apply the abbreviations for the simpler details, stich 
as An., Ca., Co., EC., Eg., h., etc. The remaining indications can be applied as the 
student progresses. 

Al., . . Ad libitum. 

An., . . Anticipation. 

App., . Appendix, or codetta. 

At., . .A tempo. 

B M., . Bell motive, or carillon. 

Ca., . . Cadenza. 

Co., . . Coda. 

Con., . . Conclusion (for sonata pieces). 

C S., . . Counter- subject, or Subtheme. 

D F., . . Dance form. 
D S., . . Dual subject (two themes com- 

EC., . . Echo. 
Eg., . . Eingang. 
Ep., . . Episode. 

G B., . Ground-base. 

H H., . Hunting-horn motive. 

1C., . . Imitation, contrary. 

IF.,.. " free. 

I R., . . " of rhythm. 

IS.,.. " strict. 

In., . . Introduction. 

Inv., . . Inversion of subject or C S. 

Iz., . . Intermezzo. 










. Motives disconnected. 
. Preliminary motive. 
. Middle theme. 

. Parenthesis. 

. Period, curtailed. 

" extended. 
. Pedal- note, or drone base. 

' ' regular. 

" united. 
. Prelude. 


. Recollection. 


. Recitative. 

Ref., . 

. Refrain. 

S A., 

. Style, Antiphonal. . 


" Harmonic. 


" Lyric. 

S T., 

. " Thematic. 

Se., . 

. Sequence. 

Str., . 

. Stretto. 

Tc., . 

. Thesis continued. 

TV., . 

. Theme varied or embellished. 

Ter., . 

. Termination. 

UP., . Uneven phrases or sections. 




Sonatinas by Steibelt ; Beethoven, op. 49 ; Reinecke, op. 136 ; Loeschhorn, op. IOI ; 
A. Krause, op. 12, I ; Alban Forster, op. 51, III; J. Handrock (in D-major). 


Templeton Strong, Cortege rustique ; Albert Biehl : Hungarian dance, op. 165 ; Spin- 
ning wheel, op. 166; Aug. Hyllested, Tempo di Polka, op. 16, II. L. Schytte, 
Enchanted Fountain, op. 63, VI. E. A. MacDowell, op. 39 (groups of I, 2, and 4 
measures). [M]. 


H. H. Huss, The Rivulet. Gustavus Johnson, Tarantella, op. 8 [M]. Isidore 
Seiss, Sonatina in />, op. 8, I, as the main theme returns, after the development, the 
four notes (a, b, c, d) should be accentuated, but without rallentando. Tschaikowsky, 
Andante from 5th symphony, transcribed for piano by Rich. Hoffman. [M D]. 


Haydn, Minuet from "Oxford" symphony (see Ex. 371). Rimsky-Korsakow, 
Romance, op. 15, II. Brahms, Hungarian dance, III. [M]. 


Mozart, Beethoven, Hummel, Dussek, Schubert, Von Weber, Rondos and Varia- 
tions. [E and M]. Ad. M. Foerster, "Eros," op. 27, I, avoided and deceptive 
cadences. [M]. John Orth, Novelette, PE. by means of avoided cadence at the 
close. [E]. 


J. S. Bach, French Suites. Eduard Schutt, op. 36, III. L. Schytte, op. 63, VI. 
Brahms, Variations of St. Anthony Choral. C. Chaminade, 2d Air de ballet. B. 
Godard, En Valsant, op. 53, VI. [M and M D]. 




Devices and Details. Mendelssohn, Duetto from " Songs without Words," SA., 
Iz., Rec. ; Nocturne from " Midsummer Night's Dream" music, PN., Rec., etc. B. 
Godard, "Pan," op. 50, SA. V. Dolmetsch, Valse lente, op. 17, EC., Eg. Louis 
Pabst, Mineaturebilder, op. 20.* E. R. Kroeger, Fantasie-Stucke. Iljinsky, Berceuse, 
op. 13, Pre., GB., etc. P. Douillet, Serenade, op. 6. E. Gillet, Babillage, Pre., Eg., 
Ter., PR. [E]. A. Rubinstein, Skip Waltz, E-flat,\i., PE.; Romance, op. 26, CS. 
M. Moszkowski, Shepherd's dance, Pre., Iz., PE. [E] ; op. 7, II, CS., etc. C. Cui, 
op. 39, VI, CS.,- PN., syncopated theme, etc. Neupert, Valse -caprice, op. 4, III, An. 
Wagner Bendel, " Walther's Prize Song," Tc., PN., pauses after the arpeggio chords. 
Th. Leschetizki, op. 9, In., Ter., etc. J. Raff, Elegy (lyric), Par., etc.; Cavatina 
(violin solo), Tc., PE., Rec. Beethoven, Polonaise from Serenade, op. 8, Rec., Str.; 
op. 14, II, in the Andante there is a Refrain of four measures. It occurs in the theme 
and in each of the variations ; Tempo di Menuetto, op. 31, III ; op. 57, first part, SA.; 
finale op. 31, II, Ca., Pas.; Rondo in C, op. 51, I, Ca., Se., Iz., Co.; Rondo from 
op. 53, Tc., Iz. Chopin, op. 31, SA. (bravura) ; Impromptu III, An.; Nocturne, 
op. 37, II, Ref. ; Valse, op. 42, Ref. ; Ballade IV, Ep. ; Rondo from op. II (E-minor 
concerto), Tc. Grieg, op. 6, II, -the real-base forms a C S. ; Dance of Elves from op. 
12, SA. Schubert- Liszt, La Serenade, echo effects. Haendel, 12 Easy Pieces (von 
Billow) Se., If., PE. Josephine Rogers, Love Song, Se., etc. Margaret R. Lang, 
Rhapsody, op. 21, Tc., Md., C S., Tempo rubato. Ph. Scharwenka, op. 38, VI (part 
II in A-minor, GB.) ; Memories, op. 72, An., Eg., Iz., If., Co., Rec.; Bagatelle in 
E-flat, Eg., etc. Ad. Henselt, Petit Valse, op. 28, I, C S.; Love Song, B-flat. 
Loeschhorn, Sonatina, op. 101, I, subject and C S. inverted; descending sequence 
crescendo in 2d theme of Scherzando ; last 19 measures. Ter. [E]. Goldner, Soli- 
tude, Eg., etc. R. H. Woodman, Romance, Eg., Par., PE. W. L. Blumenschein, 
Valse brillante, op. 23, In., Sa., Ep., PE., Ter. L. Schytte, Enchanted fountain, 
Iz., etc. J. Field, Nocturne XIII,^Iz. A. Borodin, Reverie from Petit Suite, Iz., 
Par., etc. D. Scarlatti, Harpsichord Lessons, Se., canonic imitations. S. Jadassohn, 
Canon in form of Gavotte (harmonized). Grieg, harmonized canon from op. 38. J. S. 
Bach, two- and three-part Inventions, and English Suites, canonic imitations, free, 
strict, and contrary. M. Sieveking, L'Angelus, Eg. in form of Iz., Sostenuto pedal. 
Homer Bartlett, Mazourka, op. 125, III, Ca., Pas. Brahms, Scherzo, op. 4, Iz., PE., 
etc. [D]. W. H. Sherwood, Romanza-Appassionata, op. 8, Iz., and rubato. Ben- 
del, "Am Genfer See," op. 139, I, Ep., etc. Beethoven, op. 53, the Adagio is an 
Ep. J. Field, Nocturne V, App. Grieg, Larghetto from op. 16, App., in the orches- 
tral part. E. Liebling, Spring Song, op. 33, App., etc. Bendel, op. 139, IV, Tc., 
App., Se., Ca., Pas., Co., Rec. Sgambati, Notturno, op. 20, BM. Arne Goldberg, 
op. 7, VIII, GB., etc. A. Arensky, op. 5, V, the continuous ground-base (Basso 
ostinafd) consists of two figures, ascending and descending, and these run through six 
measures before they return to the initial mensural arrangement. This constitutes a six- 
measure section, which should be duly marked whenever it recurs. 

Pedal-note and Drone Base. Grieg, op. 19, I and II. C. Cui, op. 21, IV, 
tonic and dominant pedal in form of drone base. 

Recitative. Liszt, Love dreams, I and III ; Hungarian Rhapsody VI. Ad. M. 
Foerster, op. 37, II. 

* The student is expected to discover the details when they are not here indicated. 



Coda, Termination, Recollection, Stretto. Beethoven, op. 2, III, Adagio ; op. 
8, see eleven measures marked " Calando" ; op. 14, II, coda to last variation ; Sonata 
Characteristique. Chopin, Ballades II and III ; Fantaisie Impromptu, op. 66. Karga- 
noff, op. 18, I, see " poco a poco dim. e rail." Grieg, op. 19, 1, final presto ; op. 37, 
II, Rec., and Str. ; op. 12, II. L. Gregh, Pastorale Louis XV. Th. Lack, Pendant 
la Valse, op. 73 (last seven measures). Stcherbatcheff, " Marguerite," from op. 8. 
Binding, Serenade, op. 33, IV. Sapellnikoff, Valse, op. I. 


J. S. Bach, French Suites. Haendel, Harpsichord Suites. J. de Zielinski, Bourree 
(with alternative). Arthur W. Thayer, Corrente and Bourree. R. Neimann, Gavotte, 
op. 28. 


Sgambati, Vecchio Menuetto, op. 18, II. Chas. Dennee, Tarantella (in A-minor). 
F. Thome, Tarantella, op. 43 (also Iz., PN., Ter.). R. Fiichs, Minuet from Serenade, 
op. 9. L. Schehlmann, Petit tarantella (also Se., Ca., Ter.). Nicholas Rubinstein, 
Tarantella op. 14. [D]. M. Moszkowski, Tempo di Menuetto, op. 32, I. J. Nicode, 
Tarantella op. 13, I. [M D]. Manuel Garcia, Habanera (transcribed by J. de Zielin- 
ski). L. M. Gottschalk, Souvenir de la Havane ; El Cocoye (habanera). [M and D]. 


Clavichord and Harpsichord works by F. Couperin, Corelli, D. Scarlatti, J. E. Bach, 
J. S. Bach, P. E. Bach, Haendel, and Purcell. Also modern works, such as Chopin's 
Nocturnes and Ballades; Saint-Saens' 2d Mazourka, op. 24; Seeboeck's Serenata 
Napolitan (mordents executed before the accented beat) ; J. Raff, After Sunset ; Chopin, 
op. II (see the glides in Rondo). 


A. Liadow, "Biroulki," op. 2. [M]. B, Wrangell, Chanson Naive, op. I, II. 
Wm. Mason, Mazourka Brillante, op. 49. Stcherbatcheff, Danse choral, op. 8, X. 
Tschaikowsky Album (Augener edition, 8458, a), Scherzo humoristique, groups of four 
sixteenth notes in I measure : also M D. 


Heinrich Hofmann, Hungarian dances. W. H. Sherwood, Scherzo-Caprice, op. 9 ; 
"Medea," op. 13, altered movements, punctuations, accent and pedal, indicated. A. 
M. Foerster, op. 37, I. B. O. Klein, op. 51, VI, f and \ beats remain the same, 
also pedal effect in Coda. Stcherbatcheff, Epilogue Dramatique, op. 8, XVI. 

2 9 2 



M. Moszkowski, Momento Capriccioso, op. 18, I ; Air de ballet, op. 36, V, last 
part in G-major. A. Henselt, op. 2, VI. R. Goldbeck, Petit Etude (on the motive of 
"Kelvin Grove"). Van Westerhout, Momento Capriccioso. Robert Tempest, Etude, 
op. 26, I. 


E. A. MacDowell, "Clair de Lune," from Les Orientales, op. 37, I. Emil Sjogren, 
Erotikon II (in D-fiat], F. S. Youferoff, op. I, III. B. Wrangell, Inquietude, op. 
13, II. Arthur Whiting, Idylle. G. Pierne, Venetian Serenade, op. 34 (punctuations, 
etc., indicated by the composer). Ed. Schiitt, Reverie, op. 44, V. D. M. Levett, 
Pastorale, op. 24. I. Seiss, Evening Song, op. 9. 


R. Volkmann, Song of the Hero, op. 21. Paderewski, Melodic tiree des Chants du 
Voyageur, op. 8, III, lyric-harmonic. B. O. Klein, Old Advent Hymn, Op. 55, III. 
Wickede, Fidelitas. [E]. Tschaikowsky Album (Augener edition, 8458 b), Hu- 


Chopin, Ballade III, dissonances connected ; Nocturne, op. 32, I, -ditto ; Prelude 
in -minor. 


Magic Fire Charm." Wagner Brassin. 


S. B. Mills, Album leaf in A (also Se., P R., PE., Co.). A. Grunfeld, Mazurka, op. 
14, Obligate accomp.(In., Iz., Eg., Ter., Rec.). E. Schiitt, Valse lente. 


Th. Lack, Valse Arabesque. Karganoff, Valse- caprice, op. 16. Bendel, Spinning 
wheel. E. Liebling, Menuetto-Scherzoso, op. 28 ; Valse Poetique (fictitious note- 
values, rubato, also Par., Ter., anschlag, etc.). Sjogren, Erotikon III. A. Hyllested, 
Tempo di Valse, op. 16, III. Saint-Saens, First Mazourka, op. 21. X. Scharwenka, 
Mazourka-caprice, op. 69. 




Tschaikowsky, Valse Sentimentale. Chaminade, Serenade in D y two themes com- 
bined, especially in Coda. [E]. Chopin, Scherzo, B-flat minor. B. O. Klein, " You 
and I," dual theme. [E]. J. Nicode, Ball Scenes, op. 26 (four hands), last appear- 
ance of the principal waltz theme. F. Brandeis, Andante Elegiaco, op. 71. M. Mosz- 
kowski, Gondoliera in G -minor. Wagner, " Rheingold Idyll," arranged by Carl 


A. C. Mackenzie, Evening in the Fields, Md. L. Gregh, Les Phalenes, op. 19, 
ditto. R. Niemann, Gavotte op. 28. Kjerulf- Hoffman, Scherzino. Tschaikowsky, 
Adagio Cantabile from Sextette "Souvenir de Florence," transcribed for piano by R. 


Bach-Liszt, G -minor organ fugue. Haendel, "Fire Fugue." Arthur Foote, Pre- 
lude and Fugue, op. 15. F. Busoni, Variations and Fugue, op. 22. 


Chopin, polonaise, C%-minor. Raff, Suite op. 91. Paganini-Schumann, Etude in E 
(Sternberg's edition). A. Borodin, Serenade from Petit Suite. Wrangell, Une Vision 
d' amour, op. 13, I. 


From Frescobaldi to Paradisi, and from Mozart to Richard Strauss. 


A System of Teaching 


Destined to become 

The Standard Text-Book of Musical 



of University of Pennsylvania. 

The Object kept in View 

Is how to enable the pupil to grasp, 
in the easiest, most interesting and 
comprehensible way, the mass of facts 
and rules which make up the art of 
harmony. We most earnestly invite 
all teachers and students to investi- 
gate this work, f.or it is an epoch- 
making book. To master its contents 
will place the student in possession 
of the most recent ideas in musical 
composition, by teaching him how to 
invent melodies and how to harmonize 
them correctly and effectively. 

For Class or 
Self = Instruction. 





Ifounbation /Ifoaterials 


A Piano Method Leading to Solid Musicianship 

by an Easily Graded and a Pleasingly 

Interesting Course of Study* 

EVERY measure in the book is fresh and fully up to the 
highest standard of formative materials, technically, 
musically, and pedagogically. The book is founded 
upon the best of the proved ideas of The New Teaching, \vith 
the Child-Mind ever in view. The pieces and exercises are all 
short, full of the choicest musical interest, and all have short 
and clearly defined phrases with the climax of each clearly 
marked. Every principle of touch and expressive phrasing is 
clearly stated in the ample and pungent annotations, and more 
fully illustrated in the rhythmic, melodic, and harmonic exer- 
cises, and in the best and most pleasing selections of easy, forma- 
tive music ever before embodied in a method. Special attention 
is given to the development of the inner feeling for rhythm, and 
to exact knowledge of note and rest time-values, and to ulti- 
mate skill in rapid sight-reading. Each piece and exercise 
prepares for a successful mastering of the next following 
number. There is a large number of duets for Teacher and 
Pupil, and each one illustrates an important principle in read- 
ing, time, and expression. As early impressions are lasting in 
the child's mind, every point is first given in its easiest form, 
and the annotations set it in its clearest and strongest light, 
and the subject is fully developed into free-working knowledge 
in subsequent numbers. Each page is headed with a motto, 
and these are full of striking thoughts bearing directly on 
thorough and earnest work, and they are calculated to stimu- 
late the pupil to put forth his best endeavors. The Pedal is 
taught on scientific, practical, and art principles, the pupil 
being started right regarding its use by novel and pleasing 
exerdses and pieces. The book is ideal in its contents and 
manner of presenting them, making of earnest music-study a 
pastime and pleasure of what has hitherto been too much of an 
irksome task. In engraving, printing, and paper the book is 
fully up to the standard of the works issued by this Publishing